Compare commits

...
1 Commits
Author SHA1 Message Date
ketle-manandClaude Sonnet 4.6 c5a2662bef feat: add Tagger tab + GenUI:P button + ZIT workflow fix (v0.3.38)
### Tagger tab (new)
- WD Tagger (ONNX/NCHW/NHWC), SwinV2, DeepDanbooru (.h5/TF optional), Ollama VLM
- Single subtab: drag & drop / upload, 5 output targets (GenUI:P / Prompt / Gallery / File / DB)
- Batch subtab: folder processing, Save DB / Write to File / Write .txt output options
- DB subtab: searchable SQLite, row edit, CSV export
- Backend: tagger_service.py, tagger_db_service.py, tagger_routes.py (17 endpoints)
- Security: model name validation, extension allowlist, path traversal fix in gallery_service

### Tagger single — GenUI:P button
- Appends generated tags to GenerateUI positive prompt and applies to loaded workflow immediately

### GenerateUI — ZIT/Lumina2 workflow prompt fix
- Remove ConditioningZeroOut from COND_PASSTHROUGH; shared CLIPTextEncode now correctly
  gets role "positive" instead of "unknown" in ZIT workflows
- workflow_analyzer: add lumina2 → Z-IMAGE to _CLIP_TYPE_TO_MODEL

### Gallery detail panel
- Split tab row (Info/JSON/Groups) and action row (Metadata/Load GenUI/Tagger) to separate rows
- Add Tagger button (purple) to open selected image in Tagger tab
- Fix Save to Gallery: tags now stored as array (was string, broke forEach on refresh)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-15 15:04:42 +09:00
19 changed files with 2308 additions and 9 deletions
+77
View File
@@ -1,5 +1,82 @@
# DEVLOG - ComfyUI-Workflow-Studio
## 2026-06-15: TaggerシングルGenUI:Pボタン追加+生成UIのZITワークフロープロンプト解析修正
### TaggerシングルGenUI:Pボタン(`templates/index.html`, `static/js/tagger-tab.js`, `static/js/i18n.js`)
- Single タブの結果アクション行に **GenUI:P** ボタンを追加(「プロンプトに送信」の左隣)
- クリックすると生成UIタブの `#wfm-prompt-pos-text`(ポジティブプロンプトtextarea)末尾にタグを追記し、`#wfm-prompt-pos-apply` を自動クリックしてワークフローへ即時反映
- 生成UIにワークフローが読み込まれていない場合は警告トーストを表示
- i18n: `taggerSendToGenUI` / `taggerSentToGenUI` / `taggerNoGenUI` を EN/JA/ZH に追加
- ヘルプ: `helpTagger6` を5項目に更新(GenUI:P を先頭に追加)
### 生成UIのZIT/Lumina2ワークフロープロンプト解析修正(`static/js/comfyui-workflow.js`, `py/services/workflow_analyzer.py`)
- `comfyui-workflow.js` の `COND_PASSTHROUGH` から `"ConditioningZeroOut"` を除外
- ZITワークフローでは KSampler.negative → ConditioningZeroOut → CLIPTextEncode という接続を持つが、ConditioningZeroOut は上流テキストを完全に破棄するため negative ロールを上流に伝播すべきでない
- 修正前: CLIPTextEncode が positive / negative 両ロール付与 → `getRole()` が "unknown" → Generate UI のプロンプト欄に表示されない
- 修正後: CLIPTextEncode は positive ロールのみ → 正常にポジティブプロンプトとして表示
- `workflow_analyzer.py` の `_CLIP_TYPE_TO_MODEL` に `"lumina2": "Z-IMAGE"` を追加
- CLIPLoader の `type` フィールドが `"lumina2"` のワークフローを Z-IMAGE として正確に分類
---
## 2026-06-15: Taggerバッチ.txt出力追加+Galleryタグ保存バグ修正
### Taggerバッチ.txt出力(`templates/index.html`, `static/js/tagger-tab.js`, `static/js/i18n.js`, `py/routes/tagger_routes.py`, `py/services/tagger_service.py`)
- バッチ出力オプションに **Write .txt** チェックボックスを追加
- 有効にすると処理した各画像と同一フォルダに `<ファイル名>.txt`(UTF-8)を生成、Interrogator + VLM タグのカンマ区切り文字列を書き込む
- `batch_start()` シグネチャに `write_txt: bool` パラメータを追加、ルート・フロントエンドも連動更新
- i18n: `taggerBatchWriteTxt` キーを EN/JA/ZH に追加
### Galleryタグ保存バグ修正(`static/js/tagger-tab.js`)
- `_saveToGallery()` がタグをカンマ区切り文字列のまま POST していたため、Gallery が `img.tags.forEach is not a function` エラーで更新不能になっていた
- 修正: 送信前に `tags.split(",").map(s => s.trim()).filter(Boolean)` で配列に変換
---
## 2026-06-15: Taggerタブ新規実装+ギャラリー詳細パネルUI改善+パストラバーサル修正
### Taggerタブ新規実装
#### バックエンド(`py/config.py`, `py/services/tagger_service.py`, `py/services/tagger_db_service.py`, `py/routes/tagger_routes.py`, `py/wfm.py`)
- `py/config.py`: `TAGGER_DB_FILE` / `TAGGER_SETTINGS_FILE` / `TAGGER_MODELS_DIR`(`ComfyUI/models/tagger/`)定数追加
- `TaggerService`: WD Tagger(ONNX)・SwinV2・DeepDanbooru(.h5 / TensorFlow optional)推論、Ollama VLM連携(`/api/chat`)、JPEGへのpiexif EXIF書込・PNGへのPngInfo書込・サイドカー.tags.json書込、スレッドベースバッチ処理(`batch_start` / `batch_stop` / `batch_status`)、設定の永続化
- `TaggerDbService`: SQLite(`tagger.db`)によるタグ保存・一覧・検索・更新・削除・CSV出力
- `tagger_routes.py`: `/wfm/tagger/` 配下に17エンドポイントを登録(models・predict・ollama/models・ollama/predict・batch/start/status/stop・db CRUD・write_meta・settings)、全CPU処理を `asyncio.to_thread()` でラップ
- `py/wfm.py`: `tagger_routes.setup_routes(app)` を追加
- `requirements.txt` 新規作成: `onnxruntime>=1.16`・`piexif>=1.1.3` を必須、TensorFlow をオプション(コメントアウト)として記載
#### フロントエンド(`templates/index.html`, `static/css/tagger-tab.css`, `static/js/tagger-tab.js`, `static/js/i18n.js`, `static/js/app.js`)
- `templates/index.html`: Taggerタブナビボタン追加、`<section id="wfm-tab-tagger">` を追加(Single / Batch / DBの3サブタブ構成)
- `static/css/tagger-tab.css` 新規作成: サブタブナビ・シングルレイアウト・プレビューエリア(破線ボーダー+ドラッグオーバーハイライト)・バッチ進行状況バー・DBテーブルのスタイル
- `static/js/tagger-tab.js` 新規作成:
- `initTaggerTab()`: i18n適用・サブタブ切り替え・スライダー連動・Ollamaトグル・各種イベント登録・モデル一覧ロード・設定読み込み
- `openImageInTaggerTab(img)`: Gallery→Taggerタブ遷移+画像ロード(`/wfm/gallery/image/serve` 経由でbase64変換)
- Single: ドラッグ&ドロップ・ファイルアップロード・タグ生成(WD Tagger+Ollama並行)・4出力先(Prompt送信・Gallery保存・ファイル書込・DB保存)
- Batch: バッチ開始/停止・1秒間隔ポーリングによるリアルタイム進行状況更新
- DB: 一覧表示・検索・行選択編集・保存・削除・CSV出力
- `static/js/i18n.js`: `tabTagger`・Taggerタブ関連キー約50件を EN/JA/ZH 3言語で追加
- `static/js/app.js`: `tabMap` に `tagger` 追加、`initTaggerTab()` インポート&呼び出し追加
### ギャラリー詳細パネルUI改善(`templates/index.html`, `static/css/gallery-tab.css`, `static/js/gallery-tab.js`, `static/js/app.js`)
- タブ行(Info / JSON / Groups)とアクションボタン行(Metadata / Load GenUI / Tagger)を分離: タブは `wfm-side-tab-nav`、ボタンは新設の `wfm-gallery-detail-action-row` へ移動
- `.wfm-gallery-detail-action-row` のCSSを `gallery-tab.css` に追加(`flex`・`flex-wrap`・`padding`)
- Galleryの詳細パネルに **Tagger** ボタン(`wfm-gallery-open-tagger-btn`、紫)追加: クリックで選択画像を `openImageInTaggerTab()` へ渡してTaggerタブへ遷移・画像ロード
- `.wfm-gallery-action-btn-tagger` スタイル(`gallery-tab.css`)追加
- `app.js` の `applyI18nToHtml()` でTaggerボタンのテキストを `t("tabTagger")` に設定
### セキュリティ修正: パストラバーサル対策(`py/services/tagger_service.py`, `py/services/gallery_service.py`)
- `TaggerService._validate_model_name()`: モデル名に `/` `\` `..` `\x00` が含まれる場合はロード拒否(`mdir / model_name` のパストラバーサルを防止)
- `TaggerService.write_meta_to_file()`: ユーザー指定パスを `resolve()` した上で許可拡張子(`.jpg/.jpeg/.png/.webp/.bmp/.gif`)のみ処理、それ以外はエラー返却
- `GalleryService._check_path_allowed()`: `_allowed_root is None` のときに tautology(`resolved == path.resolve()`、常にTrue)だった判定を `return False` に修正(`list_folder_tree` 呼び出し前に全パスが通る問題を解消)
### ヘルプ更新(`templates/index.html`, `static/js/i18n.js`, `static/js/app.js`)
- ヘルプナビに「Tagger Tab」追加
- Taggerタブのヘルプページ新設(9項目: モデル配置・Single操作・閾値・Ollama・出力先・Batch・DB・インストール手順)EN/JA/ZH
- Galleryタブヘルプに `helpGallery14`(Taggerボタン説明)を EN/JA/ZH で追加
- トラブルシューティングに `helpTrouble7`(Taggerモデル未表示時の対処)を EN/JA/ZH で追加
---
## 2026-06-15: 全選択ボタン追加+ギャラリーバルクバーi18n対応
### 変更内容
+22 -1
View File
@@ -17,7 +17,7 @@ A comprehensive workflow, asset management, and generation UI plugin for [ComfyU
- Built-in AI tools (translation and more)
![Workflow Studio](https://img.shields.io/badge/ComfyUI-Custom_Node-blue)
![Version](https://img.shields.io/badge/version-0.3.37-green)
![Version](https://img.shields.io/badge/version-0.3.38-green)
## Screenshots
@@ -183,6 +183,17 @@ Requires the **[comfyui-image-feeder](https://github.com/ketle-man/comfyui-image
- **Badge row** — select a badge and **+Badge** / **−Badge** to apply or remove from all selected
- **File row** — select a destination subfolder (or type a new folder name and create it) then **Move** to relocate the model file with all sidecar files; **Delete Files** (right end) permanently removes selected models and all associated files
### Tagger Tab (v0.3.38)
- **3 sub-tabs** — Single / Batch / DB for single-image tagging, folder batch processing, and tag database management
- **Model support** — WD Tagger (ONNX, NCHW/NHWC auto-detect), SwinV2 (ONNX), DeepDanbooru (.h5, requires TensorFlow); place each model in its own subfolder under `ComfyUI/models/tagger/<model-name>/` containing the `.onnx` + `selected_tags.csv` (or `.h5` + label file)
- **Threshold sliders** — General threshold (default 0.35) and Character threshold for `character:` prefix tags (default 0.85); sliders update the display value in real time
- **Ollama VLM** — optional vision model alongside WD Tagger; set API URL and select model (↻ to refresh list); results from both models are merged into one comma-separated tag string
- **Single tab** — drag & drop an image onto the preview area or click Upload; Gallery detail panel **Tagger** button opens the selected image directly here
- **Single output options** — (1) **GenUI:P** — appends tags to the GenerateUI tab's positive prompt and immediately applies to the loaded workflow; (2) **Send to Prompt** — appends tags to the Prompt tab's positive textarea; (3) **Save to Gallery** — saves tags to Gallery image metadata (requires image opened from Gallery); (4) **Write to File** — embeds tags into JPEG EXIF (`ImageDescription`) or PNG `tEXt` chunk (`Tags` key), other formats get a `.tags.json` sidecar; (5) **Save to DB** — stores in internal SQLite database
- **Batch tab** — enter a folder path, configure WD Tagger and Ollama settings; output options: **Save to DB** (default on), **Write to File** (EXIF/PNG metadata), **Write .txt** (creates `<filename>.txt` alongside each image with all tags); real-time progress bar and log; Stop cancels after the current image
- **DB tab** — searchable SQLite database of all tagged images; click a row to open the edit panel and modify WD Tags / VLM Tags; Save updates the record, Delete removes it; Export CSV downloads all records
- **Dependencies** — install into ComfyUI's embedded Python: `python_embedded\python.exe -m pip install -r requirements.txt`; for GPU inference use `onnxruntime-gpu`; TensorFlow (DeepDanbooru) is optional and commented out in `requirements.txt`
### AI TOOL Tab (v0.3.14)
- **3-pane layout** — Translation (40%) | TOOLS (40%) | Settings (20%); all panes always visible simultaneously; no sub-tab switching required
- **Translation pane** — translate text between Japanese, English, Chinese, or a custom Free language using Ollama or LM Studio; language selectors with ⇄ swap button (swaps both language selectors and text content); selections saved automatically
@@ -306,6 +317,16 @@ Click the **camera icon** (next to the W button) in ComfyUI's top bar to capture
## Changelog
### v0.3.38
- **Tagger tab** — new tab for automatic image tagging; supports WD Tagger (ONNX), SwinV2, DeepDanbooru (.h5), and Ollama VLM; 3 sub-tabs: Single (drag & drop / upload, 5 output targets), Batch (folder processing with real-time progress), DB (searchable SQLite database with CSV export)
- **Tagger single — GenUI:P button** — new button (left of "Send to Prompt") that appends generated tags to the GenerateUI positive prompt and immediately applies them to the loaded workflow; shows a warning if no workflow is loaded in GenerateUI
- **Tagger batch — Write .txt output** — new output option that creates `<filename>.txt` alongside each processed image containing all generated tags (comma-separated)
- **Gallery detail panel — Tagger button** — new action button in the Gallery detail panel opens the selected image directly in the Tagger tab for instant single-image tagging
- **Gallery detail panel layout** — tabs (Info / JSON / Groups) and action buttons (Metadata / Load GenUI / Tagger) are now on separate rows, preventing button overflow on narrow panels
- **Gallery — Save to Gallery tags fix** — tags saved from the Tagger tab were stored as a string instead of an array, causing `forEach is not a function` errors on the next Gallery refresh; now correctly stored as a string array
- **GenerateUI — ZIT/Lumina2 workflow prompt fix** — `ConditioningZeroOut` removed from the CONDITIONING passthrough set; previously it propagated the "negative" role upstream through the shared CLIPTextEncode node, marking it as "unknown" and hiding the positive prompt in the GenerateUI Input tab; ZIT (Z-Image Turbo) workflows now correctly show the positive prompt
- **Workflow analyzer — lumina2 model type** — CLIPLoader `type: "lumina2"` is now correctly classified as Z-IMAGE in the workflow type detector
### v0.3.37
- **Models tab — Select All button** — new button added to the right of "Deselect All" in the multi-select action bar; selects all models currently visible after applying the active search / filter / group filters
- **Gallery tab — Select All button** — new button added to the right of "Deselect All" in the bulk action bar; selects all images in the current folder view
+5
View File
@@ -69,3 +69,8 @@ NODE_SETS_FILE = DATA_DIR / "node_sets.json"
PROMPTS_FILE = DATA_DIR / "prompts.json"
MODEL_METADATA_FILE = DATA_DIR / "model_metadata.json"
WILDCARD_DIR = DATA_DIR / "wildcard"
# Tagger
TAGGER_DB_FILE = DATA_DIR / "tagger.db"
TAGGER_SETTINGS_FILE = DATA_DIR / "tagger_settings.json"
TAGGER_MODELS_DIR = _COMFYUI_ROOT / "models" / "tagger"
+214
View File
@@ -0,0 +1,214 @@
"""Tagger API routes."""
import asyncio
import logging
from aiohttp import web
from ..config import TAGGER_DB_FILE, TAGGER_MODELS_DIR, TAGGER_SETTINGS_FILE
from ..services.tagger_db_service import TaggerDbService
from ..services.tagger_service import TaggerService
logger = logging.getLogger(__name__)
_svc = TaggerService(TAGGER_MODELS_DIR, TAGGER_SETTINGS_FILE)
_db = TaggerDbService(TAGGER_DB_FILE)
def setup_routes(app: web.Application):
app.router.add_get("/wfm/tagger/models", handle_models)
app.router.add_post("/wfm/tagger/predict", handle_predict)
app.router.add_get("/wfm/tagger/ollama/models", handle_ollama_models)
app.router.add_post("/wfm/tagger/ollama/predict", handle_ollama_predict)
app.router.add_post("/wfm/tagger/batch/start", handle_batch_start)
app.router.add_get("/wfm/tagger/batch/status", handle_batch_status)
app.router.add_post("/wfm/tagger/batch/stop", handle_batch_stop)
app.router.add_get("/wfm/tagger/db/list", handle_db_list)
app.router.add_get("/wfm/tagger/db/search", handle_db_search)
app.router.add_put("/wfm/tagger/db/{id}", handle_db_update)
app.router.add_delete("/wfm/tagger/db/{id}", handle_db_delete)
app.router.add_get("/wfm/tagger/db/export", handle_db_export)
app.router.add_post("/wfm/tagger/db/save", handle_db_save)
app.router.add_post("/wfm/tagger/write_meta", handle_write_meta)
app.router.add_get("/wfm/tagger/settings", handle_settings_get)
app.router.add_post("/wfm/tagger/settings", handle_settings_save)
async def handle_models(request: web.Request) -> web.Response:
try:
models = await asyncio.to_thread(_svc.list_models)
return web.json_response({"models": models})
except Exception as e:
logger.error("handle_models: %s", e)
return web.json_response({"error": str(e)}, status=500)
async def handle_predict(request: web.Request) -> web.Response:
try:
body = await request.json()
b64 = body.get("image_b64", "")
model = body.get("model", "")
if not b64 or not model:
return web.json_response({"error": "image_b64 and model required"}, status=400)
result = await asyncio.to_thread(
_svc.predict, b64, model,
float(body.get("threshold", 0.35)),
float(body.get("char_threshold", 0.85)),
)
return web.json_response(result)
except Exception as e:
logger.error("handle_predict: %s", e)
return web.json_response({"error": str(e)}, status=500)
async def handle_ollama_models(request: web.Request) -> web.Response:
api_url = request.rel_url.query.get("api_url", "http://127.0.0.1:11434")
try:
models = await asyncio.to_thread(_svc.ollama_models, api_url)
return web.json_response({"models": models})
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
async def handle_ollama_predict(request: web.Request) -> web.Response:
try:
body = await request.json()
result = await asyncio.to_thread(
_svc.ollama_predict,
body.get("image_b64", ""),
body.get("api_url", "http://127.0.0.1:11434"),
body.get("model", ""),
body.get("prompt", ""),
int(body.get("max_tags", 40)),
)
return web.json_response(result)
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
async def handle_batch_start(request: web.Request) -> web.Response:
try:
body = await request.json()
result = await asyncio.to_thread(
_svc.batch_start,
body.get("folder", ""),
body.get("model", ""),
float(body.get("threshold", 0.35)),
float(body.get("char_threshold", 0.85)),
bool(body.get("use_ollama", False)),
body.get("ollama_api", "http://127.0.0.1:11434"),
body.get("ollama_model", ""),
body.get("ollama_prompt", ""),
int(body.get("ollama_max_tags", 40)),
bool(body.get("save_db", True)),
bool(body.get("write_file", False)),
bool(body.get("write_txt", False)),
_db,
)
return web.json_response(result)
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
async def handle_batch_status(request: web.Request) -> web.Response:
return web.json_response(_svc.batch_status())
async def handle_batch_stop(request: web.Request) -> web.Response:
_svc.batch_stop()
return web.json_response({"ok": True})
async def handle_db_list(request: web.Request) -> web.Response:
try:
limit = int(request.rel_url.query.get("limit", 100))
offset = int(request.rel_url.query.get("offset", 0))
rows = await asyncio.to_thread(_db.list, limit, offset)
total = await asyncio.to_thread(_db.total)
return web.json_response({"rows": rows, "total": total})
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
async def handle_db_search(request: web.Request) -> web.Response:
try:
q = request.rel_url.query.get("q", "")
rows = await asyncio.to_thread(_db.search, q)
return web.json_response({"rows": rows})
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
async def handle_db_update(request: web.Request) -> web.Response:
try:
row_id = int(request.match_info["id"])
body = await request.json()
await asyncio.to_thread(
_db.update, row_id,
body.get("interrogator_tags", ""),
body.get("vlm_tags", ""),
)
return web.json_response({"ok": True})
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
async def handle_db_delete(request: web.Request) -> web.Response:
try:
row_id = int(request.match_info["id"])
await asyncio.to_thread(_db.delete, row_id)
return web.json_response({"ok": True})
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
async def handle_db_export(request: web.Request) -> web.Response:
try:
csv_data = await asyncio.to_thread(_db.export_csv)
return web.Response(
body=csv_data.encode("utf-8"),
content_type="text/csv",
headers={"Content-Disposition": "attachment; filename=tagger_tags.csv"},
)
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
async def handle_db_save(request: web.Request) -> web.Response:
try:
body = await request.json()
row_id = await asyncio.to_thread(
_db.save,
body.get("filename", ""),
body.get("interrogator_tags", ""),
body.get("vlm_tags", ""),
)
return web.json_response({"ok": True, "id": row_id})
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
async def handle_write_meta(request: web.Request) -> web.Response:
try:
body = await request.json()
result = await asyncio.to_thread(
_svc.write_meta_to_file,
body.get("path", ""),
body.get("tags", ""),
)
return web.json_response(result)
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
async def handle_settings_get(request: web.Request) -> web.Response:
return web.json_response(_svc.load_settings())
async def handle_settings_save(request: web.Request) -> web.Response:
try:
body = await request.json()
await asyncio.to_thread(_svc.save_settings, body)
return web.json_response({"ok": True})
except Exception as e:
return web.json_response({"error": str(e)}, status=500)
+1 -2
View File
@@ -81,8 +81,7 @@ class GalleryService:
def _check_path_allowed(self, path: Path) -> bool:
"""パスが許可ルート配下かチェック(パストラバーサル防止)"""
if self._allowed_root is None:
resolved = path.resolve()
return resolved == path.resolve()
return False
try:
path.resolve().relative_to(self._allowed_root)
return True
+108
View File
@@ -0,0 +1,108 @@
"""Tagger タグ DB サービス(SQLite)。"""
import csv
import io
import logging
import sqlite3
from pathlib import Path
logger = logging.getLogger(__name__)
class TaggerDbService:
def __init__(self, db_file: Path):
self.db_file = db_file
self._init_db()
def _conn(self) -> sqlite3.Connection:
return sqlite3.connect(str(self.db_file))
def _init_db(self):
self.db_file.parent.mkdir(parents=True, exist_ok=True)
with self._conn() as conn:
conn.execute("""
CREATE TABLE IF NOT EXISTS images (
id INTEGER PRIMARY KEY AUTOINCREMENT,
filename TEXT,
interrogator_tags TEXT,
vlm_tags TEXT,
all_tags TEXT,
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
)
""")
for sql in [
"CREATE INDEX IF NOT EXISTS idx_all_tags ON images(all_tags)",
"CREATE INDEX IF NOT EXISTS idx_filename ON images(filename)",
"CREATE INDEX IF NOT EXISTS idx_created_at ON images(created_at DESC)",
]:
conn.execute(sql)
conn.commit()
def save(self, filename: str, interrogator_tags: str, vlm_tags: str) -> int:
all_tags = ", ".join(filter(None, [interrogator_tags, vlm_tags]))
with self._conn() as conn:
cur = conn.execute(
"INSERT INTO images (filename, interrogator_tags, vlm_tags, all_tags) VALUES (?, ?, ?, ?)",
(filename, interrogator_tags, vlm_tags, all_tags),
)
conn.commit()
return cur.lastrowid
def list(self, limit: int = 100, offset: int = 0) -> list:
with self._conn() as conn:
rows = conn.execute(
"SELECT id, filename, interrogator_tags, vlm_tags, all_tags, created_at "
"FROM images ORDER BY id DESC LIMIT ? OFFSET ?",
(limit, offset),
).fetchall()
return [self._row(r) for r in rows]
def total(self) -> int:
with self._conn() as conn:
return conn.execute("SELECT COUNT(*) FROM images").fetchone()[0]
def search(self, q: str) -> list:
pattern = f"%{q}%"
with self._conn() as conn:
rows = conn.execute(
"SELECT id, filename, interrogator_tags, vlm_tags, all_tags, created_at "
"FROM images WHERE all_tags LIKE ? OR filename LIKE ? ORDER BY id DESC LIMIT 200",
(pattern, pattern),
).fetchall()
return [self._row(r) for r in rows]
def update(self, row_id: int, interrogator_tags: str, vlm_tags: str):
all_tags = ", ".join(filter(None, [interrogator_tags, vlm_tags]))
with self._conn() as conn:
conn.execute(
"UPDATE images SET interrogator_tags=?, vlm_tags=?, all_tags=? WHERE id=?",
(interrogator_tags, vlm_tags, all_tags, row_id),
)
conn.commit()
def delete(self, row_id: int):
with self._conn() as conn:
conn.execute("DELETE FROM images WHERE id=?", (row_id,))
conn.commit()
def export_csv(self) -> str:
with self._conn() as conn:
rows = conn.execute(
"SELECT id, filename, interrogator_tags, vlm_tags, all_tags, created_at FROM images ORDER BY id"
).fetchall()
buf = io.StringIO()
writer = csv.writer(buf)
writer.writerow(["id", "filename", "interrogator_tags", "vlm_tags", "all_tags", "created_at"])
writer.writerows(rows)
return buf.getvalue()
@staticmethod
def _row(r) -> dict:
return {
"id": r[0],
"filename": r[1] or "",
"interrogator_tags": r[2] or "",
"vlm_tags": r[3] or "",
"all_tags": r[4] or "",
"created_at": r[5] or "",
}
+402
View File
@@ -0,0 +1,402 @@
"""WD Tagger / DeepDanbooru / Ollama VLM 推論サービス。"""
import base64
import csv
import io
import json
import logging
import tempfile
import threading
import urllib.request
from pathlib import Path
from typing import Optional
import numpy as np
from PIL import Image, ImageOps, PngImagePlugin
logger = logging.getLogger(__name__)
_model_lock = threading.Lock()
class TaggerService:
def __init__(self, models_dir: Path, settings_file: Path):
self.models_dir = models_dir
self.settings_file = settings_file
self._cache: dict = {}
self._batch_state: dict = {
"running": False, "total": 0, "done": 0, "log": [], "stop": False,
}
# ── 設定 ──────────────────────────────────────────────────
def load_settings(self) -> dict:
if self.settings_file.exists():
try:
with open(self.settings_file, "r", encoding="utf-8") as f:
return json.load(f)
except Exception:
pass
return {}
def save_settings(self, data: dict):
self.settings_file.parent.mkdir(parents=True, exist_ok=True)
with open(self.settings_file, "w", encoding="utf-8") as f:
json.dump(data, f, ensure_ascii=False, indent=2)
def _effective_models_dir(self) -> Path:
custom = self.load_settings().get("models_dir", "").strip()
if custom:
p = Path(custom)
if p.is_dir():
return p
return self.models_dir
# ── モデル一覧 ─────────────────────────────────────────────
def list_models(self) -> list:
mdir = self._effective_models_dir()
models = []
if not mdir.is_dir():
return models
for item in sorted(mdir.iterdir()):
if not item.is_dir():
continue
exts = {f.suffix.lower() for f in item.iterdir() if f.is_file()}
if ".onnx" in exts:
models.append({"name": item.name, "type": "onnx"})
elif ".h5" in exts:
models.append({"name": item.name, "type": "deepdanbooru"})
return models
# ── モデルロード ──────────────────────────────────────────
@staticmethod
def _validate_model_name(name: str) -> bool:
"""パストラバーサル防止: セパレータ・NUL・相対参照を拒否"""
return bool(name) and "/" not in name and "\\" not in name and "\x00" not in name and name not in (".", "..")
def _load_model(self, model_name: str) -> Optional[dict]:
if not self._validate_model_name(model_name):
logger.error("Invalid model name rejected: %r", model_name)
return None
if model_name in self._cache:
return self._cache[model_name]
with _model_lock:
if model_name in self._cache:
return self._cache[model_name]
mdir = self._effective_models_dir()
model_dir = mdir / model_name
if not model_dir.is_dir():
logger.error("Model dir not found: %s", model_dir)
return None
files = list(model_dir.iterdir())
onnx = next((f for f in files if f.suffix == ".onnx"), None)
h5 = next((f for f in files if f.suffix == ".h5"), None)
if onnx:
return self._load_onnx(model_name, onnx, model_dir)
if h5:
return self._load_deepdanbooru(model_name, h5, model_dir)
logger.error("No .onnx or .h5 in %s", model_dir)
return None
def _load_onnx(self, name: str, onnx_file: Path, model_dir: Path) -> Optional[dict]:
try:
import onnxruntime as ort
providers = []
try:
import torch
if torch.cuda.is_available():
providers.append("CUDAExecutionProvider")
except ImportError:
pass
providers.append("CPUExecutionProvider")
session = ort.InferenceSession(str(onnx_file), providers=providers)
labels = self._load_labels(model_dir)
data = {
"session": session,
"labels": labels,
"input_name": session.get_inputs()[0].name,
"input_shape": session.get_inputs()[0].shape,
"type": "onnx",
}
self._cache[name] = data
logger.info("ONNX loaded: %s (%d labels)", name, len(labels))
return data
except ImportError:
logger.error("onnxruntime not installed")
return None
except Exception as e:
logger.error("ONNX load error %s: %s", name, e)
return None
def _load_deepdanbooru(self, name: str, h5_file: Path, model_dir: Path) -> Optional[dict]:
try:
import tensorflow as tf
except ImportError:
logger.warning("TensorFlow not installed — DeepDanbooru unavailable")
return None
try:
model = tf.keras.models.load_model(str(h5_file))
labels = self._load_labels(model_dir)
data = {"session": model, "labels": labels, "type": "deepdanbooru"}
self._cache[name] = data
logger.info("DeepDanbooru loaded: %s (%d labels)", name, len(labels))
return data
except Exception as e:
logger.error("DeepDanbooru load error %s: %s", name, e)
return None
def _load_labels(self, model_dir: Path) -> list:
for fname in ("selected_tags.csv", "tags.csv", "tags.txt", "tags-general.txt"):
p = model_dir / fname
if not p.exists():
continue
try:
if p.suffix == ".csv":
with open(p, "r", encoding="utf-8-sig", newline="") as f:
reader = csv.reader(f)
next(reader, None)
return [row[1] for row in reader if len(row) > 1]
else:
with open(p, "r", encoding="utf-8") as f:
return [ln.strip() for ln in f if ln.strip()]
except Exception as e:
logger.warning("Label load failed %s: %s", p, e)
return []
# ── 推論(WD Tagger / DeepDanbooru) ────────────────────
def predict(self, image_b64: str, model_name: str, threshold: float = 0.35, char_threshold: float = 0.85) -> dict:
img = self._decode_b64(image_b64)
data = self._load_model(model_name)
if data is None:
return {"error": f"Model '{model_name}' could not be loaded"}
try:
if data["type"] == "deepdanbooru":
tags = self._predict_dd(img, data, threshold, char_threshold)
else:
tags = self._predict_onnx(img, data, threshold, char_threshold)
return {"tags": ", ".join(tags), "count": len(tags)}
except Exception as e:
logger.error("predict error: %s", e)
return {"error": str(e)}
def _predict_onnx(self, img: Image.Image, data: dict, thr: float, char_thr: float) -> list:
shape = data["input_shape"]
fmt = "NHWC"
size = (224, 224)
if len(shape) == 4:
if shape[1] == 3 and isinstance(shape[2], int) and shape[2] > 0:
fmt = "NCHW"
size = (int(shape[3]), int(shape[2]))
elif shape[3] == 3 and isinstance(shape[1], int) and shape[1] > 0:
fmt = "NHWC"
size = (int(shape[2]), int(shape[1]))
img = ImageOps.exif_transpose(img)
if img.mode != "RGB":
img = img.convert("RGB")
arr = np.array(img.resize(size, Image.Resampling.LANCZOS), dtype=np.float32) / 255.0
if fmt == "NCHW":
arr = arr.transpose((2, 0, 1))
arr = np.expand_dims(arr, 0)
preds = data["session"].run(None, {data["input_name"]: arr})[0][0]
return self._filter_tags(preds, data["labels"], thr, char_thr)
def _predict_dd(self, img: Image.Image, data: dict, thr: float, char_thr: float) -> list:
img = ImageOps.exif_transpose(img)
if img.mode != "RGB":
img = img.convert("RGB")
arr = np.array(img.resize((512, 512), Image.Resampling.LANCZOS), dtype=np.float32) / 255.0
arr = np.expand_dims(arr, 0)
preds = data["session"].predict(arr)[0]
return self._filter_tags(preds, data["labels"], thr, char_thr)
def _filter_tags(self, preds, labels: list, thr: float, char_thr: float) -> list:
tags = []
for i, score in enumerate(preds):
if i >= len(labels):
break
label = labels[i]
if float(score) >= (char_thr if label.startswith("character:") else thr):
tags.append(label.replace("character:", ""))
return tags
# ── Ollama VLM ────────────────────────────────────────────
def ollama_models(self, api_url: str) -> list:
try:
req = urllib.request.Request(api_url.rstrip("/") + "/api/tags")
with urllib.request.urlopen(req, timeout=5) as resp:
data = json.loads(resp.read().decode("utf-8"))
return sorted(m["name"] for m in data.get("models", []))
except Exception as e:
logger.error("ollama_models: %s", e)
return []
def ollama_predict(self, image_b64: str, api_url: str, model: str, prompt: str, max_tags: int = 40) -> dict:
payload = json.dumps({
"model": model,
"messages": [{"role": "user", "content": prompt, "images": [image_b64]}],
"stream": False,
"options": {"temperature": 0.7, "num_ctx": 4096},
}).encode("utf-8")
try:
req = urllib.request.Request(
api_url.rstrip("/") + "/api/chat",
data=payload,
headers={"Content-Type": "application/json"},
method="POST",
)
with urllib.request.urlopen(req, timeout=180) as resp:
data = json.loads(resp.read().decode("utf-8"))
content = data.get("message", {}).get("content", "")
tags = [t.strip() for t in content.split(",") if t.strip()][:max_tags]
return {"tags": ", ".join(tags), "count": len(tags)}
except Exception as e:
logger.error("ollama_predict: %s", e)
return {"error": str(e)}
# ── ファイルメタデータ書込 ────────────────────────────────
_ALLOWED_WRITE_EXTS = {".jpg", ".jpeg", ".png", ".webp", ".bmp", ".gif"}
def write_meta_to_file(self, image_path: str, tags: str) -> dict:
path = Path(image_path).resolve()
if path.suffix.lower() not in self._ALLOWED_WRITE_EXTS:
return {"error": "Unsupported file type"}
if not path.is_file():
return {"error": f"File not found: {image_path}"}
try:
import piexif
except ImportError:
piexif = None
try:
with Image.open(path) as img:
img.load()
ext = path.suffix.lower()
if ext in (".jpg", ".jpeg") and piexif:
self._write_jpeg(img, path, tags, piexif)
elif ext == ".png":
self._write_png(img, path, tags)
else:
self._write_sidecar(path, tags)
return {"ok": True}
except Exception as e:
logger.error("write_meta: %s", e)
return {"error": str(e)}
def _write_jpeg(self, img: Image.Image, path: Path, tags: str, piexif):
exif_data = img.info.get("exif")
if exif_data:
try:
exif_dict = piexif.load(exif_data)
except Exception:
exif_dict = {"0th": {}, "Exif": {}, "1st": {}, "GPS": {}, "Interop": {}}
else:
exif_dict = {"0th": {}, "Exif": {}, "1st": {}, "GPS": {}, "Interop": {}}
exif_dict["0th"][piexif.ImageIFD.ImageDescription] = tags.encode("utf-8")
exif_bytes = piexif.dump(exif_dict)
with tempfile.NamedTemporaryFile(delete=False, suffix=path.suffix) as tmp:
tmp_path = Path(tmp.name)
img.save(str(tmp_path), "JPEG", quality=95, exif=exif_bytes)
tmp_path.replace(path)
def _write_png(self, img: Image.Image, path: Path, tags: str):
meta = PngImagePlugin.PngInfo()
for k, v in img.info.items():
if isinstance(k, str) and isinstance(v, str):
meta.add_text(k, v)
meta.add_text("Tags", tags)
with tempfile.NamedTemporaryFile(delete=False, suffix=".png") as tmp:
tmp_path = Path(tmp.name)
img.save(str(tmp_path), "PNG", pnginfo=meta)
tmp_path.replace(path)
def _write_sidecar(self, path: Path, tags: str):
sidecar = path.with_suffix(".tags.json")
with open(sidecar, "w", encoding="utf-8") as f:
json.dump({"tags": tags}, f, ensure_ascii=False, indent=2)
# ── バッチ処理 ────────────────────────────────────────────
def batch_status(self) -> dict:
s = self._batch_state
return {
"running": s["running"],
"total": s["total"],
"done": s["done"],
"log": s["log"][-50:],
}
def batch_stop(self):
self._batch_state["stop"] = True
def batch_start(self, folder: str, model_name: str, threshold: float, char_threshold: float,
use_ollama: bool, ollama_api: str, ollama_model: str, ollama_prompt: str,
ollama_max_tags: int, save_db: bool, write_file: bool, write_txt: bool, db_service) -> dict:
if self._batch_state["running"]:
return {"error": "Batch already running"}
folder_path = Path(folder)
if not folder_path.is_dir():
return {"error": f"Folder not found: {folder}"}
exts = {".jpg", ".jpeg", ".png", ".webp", ".bmp", ".gif"}
files = sorted(f for f in folder_path.iterdir() if f.is_file() and f.suffix.lower() in exts)
if not files:
return {"error": "No image files found"}
self._batch_state = {
"running": True, "total": len(files), "done": 0, "log": [], "stop": False,
}
def _run():
for fpath in files:
if self._batch_state["stop"]:
self._batch_state["log"].append("Stopped.")
break
try:
with Image.open(fpath) as img:
img.load()
buf = io.BytesIO()
img.save(buf, format="PNG")
b64 = base64.b64encode(buf.getvalue()).decode("utf-8")
interrogator_tags = ""
if model_name:
res = self.predict(b64, model_name, threshold, char_threshold)
interrogator_tags = res.get("tags", "")
vlm_tags = ""
if use_ollama and ollama_model:
res2 = self.ollama_predict(b64, ollama_api, ollama_model, ollama_prompt, ollama_max_tags)
vlm_tags = res2.get("tags", "")
all_tags = ", ".join(filter(None, [interrogator_tags, vlm_tags]))
if save_db:
db_service.save(str(fpath), interrogator_tags, vlm_tags)
if write_file and all_tags:
self.write_meta_to_file(str(fpath), all_tags)
if write_txt and all_tags:
txt_path = fpath.with_suffix(".txt")
with open(txt_path, "w", encoding="utf-8") as f:
f.write(all_tags)
n = len([t for t in all_tags.split(",") if t.strip()])
self._batch_state["log"].append(f"✓ {fpath.name}: {n} tags")
except Exception as e:
self._batch_state["log"].append(f"✗ {fpath.name}: {e}")
self._batch_state["done"] += 1
self._batch_state["running"] = False
threading.Thread(target=_run, daemon=True).start()
return {"ok": True, "total": len(files)}
# ── ユーティリティ ────────────────────────────────────────
@staticmethod
def _decode_b64(b64: str) -> Image.Image:
return Image.open(io.BytesIO(base64.b64decode(b64)))
+1
View File
@@ -18,6 +18,7 @@ _CLIP_TYPE_TO_MODEL = {
"wan": "Wan",
"cosmos": "Cosmos",
"lumina": "Lumina",
"lumina2": "Z-IMAGE",
}
+2 -1
View File
@@ -61,7 +61,7 @@ class WorkflowStudio:
app.router.add_get("/wfm", serve_index_page)
# API routes
from .routes import workflow_routes, settings_routes, ollama_routes, eagle_routes, nodes_routes, prompts_routes, models_routes, gallery_routes, wildcard_routes
from .routes import workflow_routes, settings_routes, ollama_routes, eagle_routes, nodes_routes, prompts_routes, models_routes, gallery_routes, wildcard_routes, tagger_routes
workflow_routes.setup_routes(app)
settings_routes.setup_routes(app)
@@ -72,5 +72,6 @@ class WorkflowStudio:
models_routes.setup_routes(app)
gallery_routes.setup_routes(app)
wildcard_routes.setup_routes(app)
tagger_routes.setup_routes(app)
logger.info("Workflow Studio: Routes registered successfully")
+1 -1
View File
@@ -1,7 +1,7 @@
[project]
name = "comfyui-workflow-studio"
description = "A comprehensive workflow, asset management, and generation UI plugin for ComfyUI. Manage workflows, models, images, and prompts; batch generation and Image Feeder in the GenerateUI tab; library side panel with canvas drag & drop and AI tools."
version = "0.3.37"
version = "0.3.38"
license = "MIT"
requires-python = ">= 3.10"
dependencies = ["jinja2"]
+21
View File
@@ -0,0 +1,21 @@
# ComfyUI-Workflow-Studio - Tagger tab dependencies
#
# numpy and Pillow are already bundled with ComfyUI and do not need to be installed.
# torch is managed by ComfyUI and must NOT be pinned here.
#
# Install into ComfyUI's embedded Python:
# python_embeded\python.exe -m pip install -r requirements.txt
#
# For GPU-accelerated ONNX inference, install onnxruntime-gpu instead of onnxruntime:
# python_embeded\python.exe -m pip install onnxruntime-gpu
# -- Required for WD Tagger / SwinV2 ONNX inference --
onnxruntime>=1.16
# -- Required for JPEG EXIF tag writing --
piexif>=1.1.3
# -- Optional: DeepDanbooru (.h5 model) support --
# Omit if you do not use DeepDanbooru models.
# TensorFlow is large (~600MB) and conflicts with some CUDA setups.
# tensorflow>=2.10
+21 -1
View File
@@ -602,7 +602,18 @@
background: rgba(255,255,255,0.2);
}
/* ── 詳細パネル アクションボタン (タブナビ内) ────────────────── */
/* ── 詳細パネル アクションボタン行 ───────────────────────────── */
.wfm-gallery-detail-action-row {
display: flex;
gap: 6px;
padding: 4px 8px;
border-bottom: 1px solid var(--wfm-border);
flex-shrink: 0;
flex-wrap: wrap;
}
/* ── 詳細パネル アクションボタン ─────────────────────────────── */
.wfm-gallery-action-btn {
padding: 4px 10px;
@@ -634,6 +645,15 @@
border-color: var(--wfm-primary-hover);
}
.wfm-gallery-action-btn-tagger {
background: var(--wfm-accent, #7c5cfc);
border-color: var(--wfm-accent, #7c5cfc);
}
.wfm-gallery-action-btn-tagger:hover {
background: #6a4de0;
border-color: #6a4de0;
}
/* ── ファイル操作ボタン (詳細パネル) ─────────────────────────── */
.wfm-gallery-file-actions {
+294
View File
@@ -0,0 +1,294 @@
/* ============================================================
Tagger Tab Styles
============================================================ */
/* ── サブタブナビ ────────────────────────────────────────────── */
.wfm-tagger-subtab-nav {
display: flex;
gap: 0;
border-bottom: 1px solid var(--wfm-border);
background: var(--wfm-bg-secondary);
flex-shrink: 0;
}
.wfm-tagger-subtab-btn {
padding: 8px 20px;
font-size: 13px;
font-weight: 500;
background: transparent;
border: none;
border-bottom: 2px solid transparent;
color: var(--wfm-text-secondary);
cursor: pointer;
transition: color 0.15s, border-color 0.15s;
white-space: nowrap;
}
.wfm-tagger-subtab-btn:hover {
color: var(--wfm-text);
}
.wfm-tagger-subtab-btn.active {
color: var(--wfm-primary);
border-bottom-color: var(--wfm-primary);
}
/* ── サブタブ共通 ────────────────────────────────────────────── */
.wfm-tagger-subtab {
flex: 1;
overflow-y: auto;
padding: 12px;
}
/* ── セクション ─────────────────────────────────────────────── */
.wfm-tagger-section {
background: var(--wfm-surface);
border: 1px solid var(--wfm-border);
border-radius: var(--wfm-radius);
padding: 10px 12px;
margin-bottom: 10px;
}
.wfm-tagger-section-header {
font-size: 12px;
font-weight: 600;
color: var(--wfm-text-secondary);
text-transform: uppercase;
letter-spacing: 0.05em;
margin-bottom: 8px;
}
/* ── フォームレイアウト ─────────────────────────────────────── */
.wfm-tagger-form-row {
display: flex;
align-items: center;
gap: 8px;
margin-bottom: 6px;
}
.wfm-tagger-form-row:last-child {
margin-bottom: 0;
}
.wfm-tagger-label {
font-size: 12px;
color: var(--wfm-text-secondary);
min-width: 80px;
white-space: nowrap;
}
.wfm-tagger-select {
flex: 1;
min-width: 0;
}
/* ── スライダー ─────────────────────────────────────────────── */
.wfm-tagger-slider {
flex: 1;
min-width: 0;
accent-color: var(--wfm-primary);
cursor: pointer;
}
.wfm-tagger-slider-val {
font-size: 12px;
color: var(--wfm-text-secondary);
min-width: 36px;
text-align: right;
}
/* ── Single サブタブ ─────────────────────────────────────────── */
.wfm-tagger-single-layout {
display: flex;
gap: 12px;
height: calc(100vh - var(--wfm-header-height, 96px) - 80px);
min-height: 0;
}
.wfm-tagger-image-pane {
width: 300px;
min-width: 240px;
display: flex;
flex-direction: column;
gap: 8px;
flex-shrink: 0;
}
.wfm-tagger-image-sources {
display: flex;
gap: 6px;
flex-wrap: wrap;
}
.wfm-tagger-preview-wrap {
flex: 1;
min-height: 0;
background: var(--wfm-surface);
border: 2px dashed var(--wfm-border);
border-radius: var(--wfm-radius);
display: flex;
align-items: center;
justify-content: center;
overflow: hidden;
position: relative;
transition: border-color 0.15s, background 0.15s;
cursor: pointer;
}
.wfm-tagger-preview-wrap.drag-over {
border-color: var(--wfm-primary);
background: rgba(74, 158, 255, 0.08);
}
.wfm-tagger-image-name {
font-size: 11px;
color: var(--wfm-text-secondary);
text-align: center;
word-break: break-all;
min-height: 16px;
}
.wfm-tagger-controls-pane {
flex: 1;
min-width: 0;
overflow-y: auto;
display: flex;
flex-direction: column;
gap: 0;
}
.wfm-tagger-result-actions {
display: flex;
gap: 6px;
margin-top: 6px;
flex-wrap: wrap;
}
/* ── Batch サブタブ ──────────────────────────────────────────── */
.wfm-tagger-batch-layout {
max-width: 900px;
}
.wfm-tagger-batch-cols {
display: grid;
grid-template-columns: 1fr 1fr;
gap: 10px;
}
.wfm-tagger-progress-bar-wrap {
height: 6px;
background: var(--wfm-border);
border-radius: 3px;
overflow: hidden;
}
.wfm-tagger-progress-bar {
height: 100%;
background: var(--wfm-primary);
border-radius: 3px;
transition: width 0.3s ease;
}
/* ── DB サブタブ ─────────────────────────────────────────────── */
.wfm-tagger-db-layout {
display: flex;
flex-direction: column;
height: calc(100vh - var(--wfm-header-height, 96px) - 80px);
min-height: 0;
}
.wfm-tagger-db-toolbar {
display: flex;
gap: 6px;
align-items: center;
margin-bottom: 10px;
flex-shrink: 0;
}
.wfm-tagger-db-table-wrap {
flex: 1;
overflow: auto;
border: 1px solid var(--wfm-border);
border-radius: var(--wfm-radius);
min-height: 0;
}
.wfm-tagger-db-table {
width: 100%;
border-collapse: collapse;
font-size: 12px;
}
.wfm-tagger-db-table th,
.wfm-tagger-db-table td {
padding: 6px 10px;
text-align: left;
border-bottom: 1px solid var(--wfm-border);
white-space: nowrap;
overflow: hidden;
text-overflow: ellipsis;
max-width: 220px;
}
.wfm-tagger-db-table th {
background: var(--wfm-bg-secondary);
color: var(--wfm-text-secondary);
font-weight: 600;
text-transform: uppercase;
font-size: 11px;
letter-spacing: 0.04em;
position: sticky;
top: 0;
z-index: 1;
}
.wfm-tagger-db-table tbody tr {
cursor: pointer;
transition: background 0.1s;
}
.wfm-tagger-db-table tbody tr:hover {
background: var(--wfm-surface-hover);
}
.wfm-tagger-db-table tbody tr.selected {
background: rgba(74, 158, 255, 0.15);
}
.wfm-tagger-db-edit-panel {
flex-shrink: 0;
margin-top: 10px;
background: var(--wfm-surface);
border: 1px solid var(--wfm-border);
border-radius: var(--wfm-radius);
padding: 10px 12px;
}
/* ── ユーティリティ ─────────────────────────────────────────── */
#wfm-tab-tagger {
display: none;
flex-direction: column;
height: calc(100vh - var(--wfm-header-height, 56px));
overflow: hidden;
}
#wfm-tab-tagger.active {
display: flex;
}
.wfm-placeholder {
color: var(--wfm-text-secondary);
font-size: 13px;
text-align: center;
padding: 20px;
pointer-events: none;
}
+13 -2
View File
@@ -91,7 +91,7 @@ function initTabs() {
function applyI18nToHtml() {
// Tab labels
const tabMap = { workflow: "tabWorkflow", nodes: "tabNodes", models: "tabModels", generate: "tabGenerate", prompt: "tabPrompt", metadata: "tabMetadata", gallery: "tabGallery", settings: "tabSettings", help: "tabHelp", ai: "tabAi" };
const tabMap = { workflow: "tabWorkflow", nodes: "tabNodes", models: "tabModels", generate: "tabGenerate", prompt: "tabPrompt", metadata: "tabMetadata", gallery: "tabGallery", settings: "tabSettings", help: "tabHelp", ai: "tabAi", tagger: "tabTagger" };
document.querySelectorAll(".wfm-tab").forEach((tab) => {
const key = tabMap[tab.dataset.tab];
if (key) tab.textContent = t(key);
@@ -215,6 +215,8 @@ function applyI18nToHtml() {
if (galleryBulkMove) galleryBulkMove.textContent = t("galleryBulkMoveTo");
const galleryBulkDelete = document.getElementById("wfm-gallery-bulk-delete");
if (galleryBulkDelete) galleryBulkDelete.textContent = t("galleryBulkDelete");
const galleryOpenTaggerBtn = document.getElementById("wfm-gallery-open-tagger-btn");
if (galleryOpenTaggerBtn) galleryOpenTaggerBtn.textContent = t("tabTagger");
// Prompt tab
const assistantHeader = document.querySelector(".wfm-prompt-split-left .wfm-prompt-split-header");
@@ -384,7 +386,13 @@ function applyI18nToHtml() {
"wfm-help-gallery-7": "helpGallery7", "wfm-help-gallery-8": "helpGallery8",
"wfm-help-gallery-9": "helpGallery9", "wfm-help-gallery-10": "helpGallery10",
"wfm-help-gallery-11": "helpGallery11", "wfm-help-gallery-12": "helpGallery12",
"wfm-help-gallery-13": "helpGallery13",
"wfm-help-gallery-13": "helpGallery13", "wfm-help-gallery-14": "helpGallery14",
"wfm-help-tagger-title": "helpTaggerTitle",
"wfm-help-tagger-1": "helpTagger1", "wfm-help-tagger-2": "helpTagger2",
"wfm-help-tagger-3": "helpTagger3", "wfm-help-tagger-4": "helpTagger4",
"wfm-help-tagger-5": "helpTagger5", "wfm-help-tagger-6": "helpTagger6",
"wfm-help-tagger-7": "helpTagger7", "wfm-help-tagger-8": "helpTagger8",
"wfm-help-tagger-9": "helpTagger9",
"wfm-help-sidepanel-title": "helpSidepanelTitle",
"wfm-help-sidepanel-1": "helpSidepanel1", "wfm-help-sidepanel-2": "helpSidepanel2",
"wfm-help-sidepanel-3": "helpSidepanel3", "wfm-help-sidepanel-4": "helpSidepanel4",
@@ -405,6 +413,7 @@ function applyI18nToHtml() {
"wfm-help-trouble-1": "helpTrouble1", "wfm-help-trouble-2": "helpTrouble2",
"wfm-help-trouble-3": "helpTrouble3", "wfm-help-trouble-4": "helpTrouble4",
"wfm-help-trouble-5": "helpTrouble5", "wfm-help-trouble-6": "helpTrouble6",
"wfm-help-trouble-7": "helpTrouble7",
"wfm-help-support-title": "helpSupportTitle",
"wfm-help-support-desc": "helpSupportDesc",
"wfm-help-github-desc": "helpGithubDesc",
@@ -457,6 +466,7 @@ import { initModelsTab } from "./models-tab.js";
import { initGalleryTab } from "./gallery-tab.js";
import { initMetadataTab } from "./metadata-tab.js";
import { initAiTab } from "./ai-tab.js";
import { initTaggerTab } from "./tagger-tab.js";
// Apply saved theme immediately to prevent flash of default theme
applyTheme(getSavedTheme());
@@ -494,6 +504,7 @@ document.addEventListener("DOMContentLoaded", () => {
initMetadataTab();
initGalleryTab();
initAiTab();
initTaggerTab();
console.log("Workflow Studio: App initialized");
});
+4 -1
View File
@@ -520,7 +520,10 @@ export const comfyWorkflow = {
// Nodes that pass CONDITIONING through without changing positive/negative role
const COND_PASSTHROUGH = new Set([
"ConditioningCombine", "ConditioningConcat", "ConditioningAverage",
"ConditioningZeroOut", "ConditioningSetTimestepRange",
"ConditioningSetTimestepRange",
// ConditioningZeroOut is intentionally excluded: it discards upstream text entirely
// (used as a no-text negative in ZIT/Lumina2 workflows). Propagating the negative
// role upstream would incorrectly mark the shared positive CLIPTextEncode as "unknown".
"ControlNetApply", "ControlNetApplyAdvanced",
"IPAdapterApply", "IPAdapterApplyFaceID",
"StyleModelApply",
+9
View File
@@ -1300,6 +1300,15 @@ function bindEvents() {
openImageInMetadataTab(state.selectedImage);
});
document.getElementById("wfm-gallery-open-tagger-btn")?.addEventListener("click", async () => {
if (!state.selectedImage) {
showToast(t("gallerySelectImageFirst"), "error");
return;
}
const { openImageInTaggerTab } = await import("./tagger-tab.js");
openImageInTaggerTab(state.selectedImage);
});
// Load in GenerateUI ボタン: 埋め込みワークフローをGenerateUIタブに読み込む
document.getElementById("wfm-gallery-load-genui-btn")?.addEventListener("click", async () => {
if (!state.selectedImage) {
+201
View File
@@ -468,6 +468,58 @@ const LANGUAGES = {
galleryBulkDelete: "Delete",
tabAi: "AI TOOL",
tabTagger: "Tagger",
// -- Tagger Tab --
taggerSingleTab: "Single",
taggerBatchTab: "Batch",
taggerDbTab: "DB",
taggerFromGallery: "From Gallery",
taggerUpload: "Upload",
taggerModel: "Model",
taggerThreshold: "Threshold",
taggerCharThreshold: "Char Thr.",
taggerGenerateTags: "Generate Tags",
taggerOllamaEnable: "Ollama VLM",
taggerOllamaUrl: "API URL",
taggerOllamaModel: "Model",
taggerOllamaPrompt: "Prompt",
taggerOllamaMaxTags: "Max Tags",
taggerResult: "Result Tags",
taggerSendToGenUI: "GenUI:P",
taggerSendToPrompt: "Send to Prompt",
taggerSaveToGallery: "Save to Gallery",
taggerWriteToFile: "Write to File",
taggerBatchWriteTxt: "Write .txt",
taggerSaveToDB: "Save to DB",
taggerBatchFolder: "Image Folder",
taggerBatchOutput: "Output Options",
taggerBatchStart: "Start Batch",
taggerBatchStop: "Stop",
taggerDbSearch: "Search",
taggerDbShowAll: "Show All",
taggerDbExport: "Export CSV",
taggerDbFile: "Filename",
taggerDbInterrogator: "WD Tags",
taggerDbVlm: "VLM Tags",
taggerDbDate: "Date",
taggerDbSave: "Save",
taggerDbDelete: "Delete",
taggerDbSaved: "Saved.",
taggerDbDeleted: "Deleted.",
taggerDbEmpty: "No records.",
taggerNoImage: "No image loaded",
taggerNoModel: "-- No model --",
taggerNoTags: "No tags generated",
taggerDone: "Done",
taggerSentToGenUI: "Sent to GenerateUI positive prompt",
taggerNoGenUI: "GenerateUI has no workflow loaded",
taggerSentToPrompt: "Sent to Prompt tab",
taggerSavedToGallery: "Saved to Gallery",
taggerNoGalleryImage: "No image path available",
taggerWrittenToFile: "Written to file",
taggerSavedToDB: "Saved to DB",
taggerSelectGalleryImage: "Select an image in the Gallery tab",
// -- AI Tab --
aiSubTabTranslate: "Translate",
@@ -607,6 +659,20 @@ const LANGUAGES = {
helpGallery11: "Supports PNG (tEXt/iTXt), JPEG, WebP, GIF; workflow extraction reads ComfyUI prompt and workflow keys embedded in PNG metadata",
helpGallery12: "Metadata button (green, in the detail panel tab bar): opens the selected image in the Metadata tab for full embedded-data inspection; Alt+click on any image thumbnail has the same effect",
helpGallery13: "Load GenUI button (blue, in the detail panel tab bar): loads the embedded ComfyUI workflow from the selected image directly into the GenerateUI tab and switches to it; shows a warning toast if no workflow is embedded or the format is not supported",
helpGallery14: "Tagger button (purple, in the detail panel action row): opens the selected image directly in the Tagger tab's Single mode for instant tagging",
// -- Tagger Tab --
helpTaggerTitle: "Tagger Tab",
helpTagger1: "Automatic image tagging using WD Tagger (ONNX), SwinV2, DeepDanbooru (.h5), and Ollama VLM — 3 sub-tabs: Single / Batch / DB",
helpTagger2: "Model placement: put each model in its own subfolder under ComfyUI/models/tagger/<model-name>/; the folder must contain a .onnx file + selected_tags.csv (or .h5 + label file for DeepDanbooru)",
helpTagger3: "Single tab: drag & drop an image onto the preview area, or click Upload to pick a file; the Tagger button in the Gallery detail panel opens the selected image directly here",
helpTagger4: "Threshold slider: minimum confidence score for general tags (default 0.35); Character Threshold: separate threshold for character: prefix tags (default 0.85)",
helpTagger5: "Ollama VLM: enable the checkbox to run a vision model alongside WD Tagger; set API URL, select model (click ↻ to refresh), and customize the prompt; results from both models are merged",
helpTagger6: "Result tags can be sent to: (1) GenUI:P — appended to the GenerateUI tab's positive prompt textarea and applied to the loaded workflow immediately; (2) Send to Prompt — appended to the Prompt tab's positive textarea; (3) Save to Gallery — saved to Gallery image metadata (requires image opened from Gallery); (4) Write to File — embedded into JPEG EXIF (ImageDescription) or PNG tEXt (Tags key); (5) Save to DB — stored in the internal SQLite database",
helpTagger7: "Batch tab: enter a folder path, configure WD Tagger and Ollama settings; output options: Save to DB (default on), Write to File (EXIF/PNG metadata), Write .txt (creates <filename>.txt alongside each image); click Start Batch to process all images in the folder; progress bar and log are updated in real time; Stop cancels after the current image",
helpTagger8: "DB tab: searchable database of all tagged images; click a row to open the edit panel and modify WD Tags / VLM Tags; Save updates the record, Delete removes it; Export CSV downloads all records as a CSV file",
helpTagger9: "Install dependencies into ComfyUI's embedded Python: python_embeded\\python.exe -m pip install -r requirements.txt; for GPU ONNX inference install onnxruntime-gpu instead of onnxruntime; TensorFlow (for DeepDanbooru) is commented out by default due to size",
helpSidepanelTitle: "Workflow Studio Library (ComfyUI Side Panel)",
helpSidepanel1: "Open from the Workflow Studio Library button in the ComfyUI toolbar",
helpSidepanel2: "Workflows tab: All / ★ Favorites / Groups sub-tabs; ★ shown for favorites in All view",
@@ -643,6 +709,7 @@ const LANGUAGES = {
helpTrouble4: "Image generation fails: Check if specified models exist in ComfyUI, verify connection, try restarting",
helpTrouble5: "LoRA Stack fails with 'No such file or directory' error: The LoRA may not be recognized by comfyui-lora-manager (known issue). Try a different LoRA or check if the file appears in Lora Manager's own UI",
helpTrouble6: "Model loading error after moving or deleting models: ComfyUI caches file paths and may still reference old locations. Perform a Hard Reset from the ComfyUI menu (Queue → Hard Reset) or restart ComfyUI to clear the cache.",
helpTrouble7: "Tagger tab shows no models: verify the model folder exists at ComfyUI/models/tagger/<model-name>/ and contains a .onnx or .h5 file; if onnxruntime is not installed, run: python_embeded\\python.exe -m pip install onnxruntime",
helpSupportTitle: "Support",
helpSupportDesc: "If you find this plugin useful, please consider supporting its development!",
helpGithubDesc: "Report issues, request features, and contribute",
@@ -1248,6 +1315,58 @@ const LANGUAGES = {
galleryBulkDelete: "削除",
tabAi: "AI TOOL",
tabTagger: "Tagger",
// -- Tagger Tab --
taggerSingleTab: "シングル",
taggerBatchTab: "バッチ",
taggerDbTab: "DB",
taggerFromGallery: "Galleryから",
taggerUpload: "アップロード",
taggerModel: "モデル",
taggerThreshold: "閾値",
taggerCharThreshold: "キャラ閾値",
taggerGenerateTags: "タグ生成",
taggerOllamaEnable: "Ollama VLM",
taggerOllamaUrl: "API URL",
taggerOllamaModel: "モデル",
taggerOllamaPrompt: "プロンプト",
taggerOllamaMaxTags: "最大タグ数",
taggerResult: "生成タグ",
taggerSendToGenUI: "GenUI:P",
taggerSendToPrompt: "プロンプトに送信",
taggerSaveToGallery: "Galleryに保存",
taggerWriteToFile: "ファイルに書込",
taggerBatchWriteTxt: ".txt出力",
taggerSaveToDB: "DBに保存",
taggerBatchFolder: "画像フォルダ",
taggerBatchOutput: "出力オプション",
taggerBatchStart: "バッチ開始",
taggerBatchStop: "停止",
taggerDbSearch: "検索",
taggerDbShowAll: "全表示",
taggerDbExport: "CSV出力",
taggerDbFile: "ファイル名",
taggerDbInterrogator: "WDタグ",
taggerDbVlm: "VLMタグ",
taggerDbDate: "日付",
taggerDbSave: "保存",
taggerDbDelete: "削除",
taggerDbSaved: "保存しました",
taggerDbDeleted: "削除しました",
taggerDbEmpty: "レコードなし",
taggerNoImage: "画像が読み込まれていません",
taggerNoModel: "-- モデルなし --",
taggerNoTags: "タグが生成されていません",
taggerDone: "完了",
taggerSentToGenUI: "生成UIのポジティブプロンプトに送信しました",
taggerNoGenUI: "生成UIにワークフローが読み込まれていません",
taggerSentToPrompt: "プロンプトタブに送信しました",
taggerSavedToGallery: "Galleryに保存しました",
taggerNoGalleryImage: "画像パスがありません",
taggerWrittenToFile: "ファイルに書き込みました",
taggerSavedToDB: "DBに保存しました",
taggerSelectGalleryImage: "Galleryタブで画像を選択してください",
// -- AI Tab --
aiSubTabTranslate: "翻訳",
@@ -1387,6 +1506,20 @@ const LANGUAGES = {
helpGallery11: "PNG (tEXt/iTXt)、JPEG、WebP、GIF対応;PNG形式はComfyUIの埋め込みJSON(prompt / workflow キー)を読み込んでワークフロー抽出が可能",
helpGallery12: "Metadataボタン(緑、詳細パネルのタブバー内):選択画像をMetadataタブで開いて埋め込みデータを詳細確認。画像サムネイルのAlt+クリックでも同じ操作が可能",
helpGallery13: "Load GenUIボタン(青、詳細パネルのタブバー内):選択画像に埋め込まれたComfyUIワークフローをGenerateUIタブへ直接読み込んで切り替え。ワークフローが埋め込まれていない場合や非対応形式の場合は警告トーストを表示",
helpGallery14: "Taggerボタン(紫、詳細パネルのアクション行):選択画像をTaggerタブのSingleモードで直接開き、即座にタグ付けを開始できる",
// -- Tagger Tab --
helpTaggerTitle: "Taggerタブ",
helpTagger1: "WD Tagger(ONNX)・SwinV2・DeepDanbooru(.h5)・Ollama VLMによる自動タグ付け — サブタブ: Single / Batch / DB",
helpTagger2: "モデル配置:ComfyUI/models/tagger/<モデル名>/ 以下に各モデルのフォルダを作成し、.onnx + selected_tags.csv(DeepDanbooruは .h5 + ラベルファイル)を格納",
helpTagger3: "Singleタブ:プレビューエリアへ画像をドラッグ&ドロップ、またはUploadボタンでファイルを選択;GalleryタブのTaggerボタンで選択画像を直接開くことも可能",
helpTagger4: "閾値スライダー:一般タグの信頼スコア下限(デフォルト0.35);キャラ閾値:character: プレフィックス付きタグ専用の閾値(デフォルト0.85)",
helpTagger5: "Ollama VLM:チェックボックスをオンにするとWD Taggerと並行してビジョンモデルを実行;API URLとモデルを設定(↻で一覧更新)、プロンプトを編集可能;両モデルの結果はマージされる",
helpTagger6: "生成タグの送信先:(1) GenUI:P — 生成UIタブのポジティブプロンプトに追記し、読み込み中のワークフローへ即時適用;(2) プロンプトに送信 — Promptタブの正プロンプトテキストエリアに追記;(3) Galleryに保存 — ギャラリーの画像メタデータに保存(Galleryから開いた画像のみ有効);(4) ファイルに書込 — JPEGはEXIF(ImageDescription)、PNGはtEXt(Tagsキー)に書き込み;(5) DBに保存 — 内蔵SQLiteデータベースに保存",
helpTagger7: "Batchタブ:処理対象フォルダパスを入力してWD Tagger・Ollamaの設定を行い、出力オプション(DBに保存〈デフォルトON〉・ファイルに書込〈EXIF/PNGメタデータ〉・.txt出力〈画像と同フォルダに <ファイル名>.txt を生成〉)を選択;バッチ開始で指定フォルダ内の全画像を処理;進行状況バーとログがリアルタイム更新;停止ボタンは現在の画像処理完了後にキャンセル",
helpTagger8: "DBタブ:タグ付けした画像の検索可能データベース;行をクリックすると編集パネルでWDタグ・VLMタグを編集可能;保存で更新、削除でレコード削除;CSV出力で全レコードをダウンロード",
helpTagger9: "依存パッケージのインストール:python_embeded\\python.exe -m pip install -r requirements.txt;GPU推論にはonnxruntimeの代わりにonnxruntime-gpuをインストール;TensorFlow(DeepDanbooru用)はサイズ大のためデフォルトではコメントアウト",
helpSidepanelTitle: "Workflow Studio Library(ComfyUIサイドパネル)",
helpSidepanel1: "ComfyUIツールバーのWorkflow Studio Libraryボタンから開く",
helpSidepanel2: "Workflowsタブ:All / ★ Favorites / Groupsサブタブ、Allビューでお気に入りに★表示",
@@ -1423,6 +1556,7 @@ const LANGUAGES = {
helpTrouble4: "画像生成が失敗する:指定モデルの存在確認、ComfyUI接続確認、再起動",
helpTrouble5: "LoRAスタック実行時に「No such file or directory」エラー:comfyui-lora-managerが該当LoRAを認識していない可能性があります(既知の問題)。別のLoRAを試すか、Lora Manager自体のUIで該当ファイルが表示されるか確認してください",
helpTrouble6: "モデルの移動・削除後にモデル読み込みエラーが発生する:ComfyUIがファイルパスをキャッシュしており、古い場所を参照することがあります。ComfyUIメニューのHard Reset(Queue → Hard Reset)またはComfyUIを再起動してキャッシュをクリアしてください。",
helpTrouble7: "Taggerタブにモデルが表示されない:ComfyUI/models/tagger/<モデル名>/ フォルダの存在と .onnx または .h5 ファイルの有無を確認;onnxruntimeが未インストールの場合は python_embeded\\python.exe -m pip install onnxruntime を実行",
helpSupportTitle: "サポート",
helpSupportDesc: "このプラグインが役に立ったら、開発を支援していただけると嬉しいです!",
helpGithubDesc: "バグ報告、機能リクエスト、コントリビュート",
@@ -2028,6 +2162,58 @@ const LANGUAGES = {
galleryBulkDelete: "删除",
tabAi: "AI TOOL",
tabTagger: "Tagger",
// -- Tagger Tab --
taggerSingleTab: "单张",
taggerBatchTab: "批量",
taggerDbTab: "DB",
taggerFromGallery: "从图库",
taggerUpload: "上传",
taggerModel: "模型",
taggerThreshold: "阈值",
taggerCharThreshold: "角色阈值",
taggerGenerateTags: "生成标签",
taggerOllamaEnable: "Ollama VLM",
taggerOllamaUrl: "API URL",
taggerOllamaModel: "模型",
taggerOllamaPrompt: "提示词",
taggerOllamaMaxTags: "最大标签数",
taggerResult: "生成标签",
taggerSendToGenUI: "GenUI:P",
taggerSendToPrompt: "发送到提示词",
taggerSaveToGallery: "保存到图库",
taggerWriteToFile: "写入文件",
taggerBatchWriteTxt: "输出.txt",
taggerSaveToDB: "保存到DB",
taggerBatchFolder: "图像文件夹",
taggerBatchOutput: "输出选项",
taggerBatchStart: "开始批处理",
taggerBatchStop: "停止",
taggerDbSearch: "搜索",
taggerDbShowAll: "显示全部",
taggerDbExport: "导出CSV",
taggerDbFile: "文件名",
taggerDbInterrogator: "WD标签",
taggerDbVlm: "VLM标签",
taggerDbDate: "日期",
taggerDbSave: "保存",
taggerDbDelete: "删除",
taggerDbSaved: "已保存",
taggerDbDeleted: "已删除",
taggerDbEmpty: "无记录",
taggerNoImage: "未加载图像",
taggerNoModel: "-- 无模型 --",
taggerNoTags: "未生成标签",
taggerDone: "完成",
taggerSentToGenUI: "已发送到生成UI正向提示词",
taggerNoGenUI: "生成UI未加载工作流",
taggerSentToPrompt: "已发送到提示词标签",
taggerSavedToGallery: "已保存到图库",
taggerNoGalleryImage: "无图像路径",
taggerWrittenToFile: "已写入文件",
taggerSavedToDB: "已保存到DB",
taggerSelectGalleryImage: "请在图库标签中选择图像",
// -- AI Tab --
aiSubTabTranslate: "翻译",
@@ -2167,6 +2353,20 @@ const LANGUAGES = {
helpGallery11: "支持PNG(tEXt/iTXt)、JPEG、WebP、GIF;PNG格式读取ComfyUI嵌入的JSON(prompt / workflow键)提取工作流",
helpGallery12: "Metadata按钮(绿色,位于详情面板标签栏):在Metadata标签中打开所选图像以全面查看嵌入数据;Alt+点击任意图像缩略图效果相同",
helpGallery13: "Load GenUI按钮(蓝色,位于详情面板标签栏):将所选图像中嵌入的ComfyUI工作流直接加载到GenerateUI标签并切换过去;如果没有嵌入工作流或格式不受支持,则显示警告提示",
helpGallery14: "Tagger按钮(紫色,位于详情面板操作行):将所选图像直接在Tagger标签的Single模式中打开,即可立即进行标注",
// -- Tagger Tab --
helpTaggerTitle: "Tagger标签",
helpTagger1: "使用WD Tagger(ONNX)、SwinV2、DeepDanbooru(.h5)和Ollama VLM进行自动图像标注 — 子标签:Single / Batch / DB",
helpTagger2: "模型放置:在ComfyUI/models/tagger/<模型名>/ 下为每个模型创建单独子文件夹,必须包含 .onnx + selected_tags.csv(DeepDanbooru为 .h5 + 标签文件)",
helpTagger3: "Single标签:将图像拖放到预览区域,或点击Upload选择文件;图库详情面板的Tagger按钮可直接在此处打开所选图像",
helpTagger4: "阈值滑块:一般标签的最低置信度(默认0.35);角色阈值:character: 前缀标签的独立阈值(默认0.85)",
helpTagger5: "Ollama VLM:勾选复选框可与WD Tagger并行运行视觉模型;设置API URL、选择模型(点击↻刷新列表)并自定义提示词;两个模型的结果将合并",
helpTagger6: "生成标签的发送目标:(1) GenUI:P — 追加到生成UI标签的正向提示词并立即应用到已加载工作流;(2) 发送到提示词 — 追加到提示词标签的正向文本区;(3) 保存到图库 — 保存到图库图像元数据(仅限从图库打开的图像);(4) 写入文件 — JPEG写入EXIF(ImageDescription)、PNG写入tEXt(Tags键);(5) 保存到DB — 存入内置SQLite数据库",
helpTagger7: "Batch标签:输入目标文件夹路径并配置WD Tagger和Ollama设置,选择输出选项(保存DB〈默认开启〉、写入文件〈EXIF/PNG元数据〉、输出.txt〈在图像同目录生成 <文件名>.txt〉);开始批处理后将处理文件夹内所有图像;进度条和日志实时更新;停止按钮在当前图像处理完成后取消",
helpTagger8: "DB标签:所有已标注图像的可搜索数据库;点击行可在编辑面板中修改WD标签/VLM标签;保存更新记录,删除移除记录;导出CSV可下载所有记录",
helpTagger9: "安装依赖包:python_embeded\\python.exe -m pip install -r requirements.txt;GPU推理请安装onnxruntime-gpu替代onnxruntime;TensorFlow(用于DeepDanbooru)因体积较大默认已注释",
helpSidepanelTitle: "Workflow Studio Library(ComfyUI侧面板)",
helpSidepanel1: "从ComfyUI工具栏的Workflow Studio Library按钮打开",
helpSidepanel2: "Workflows标签:All / ★ Favorites / Groups子标签,All视图中收藏项显示★",
@@ -2203,6 +2403,7 @@ const LANGUAGES = {
helpTrouble4: "图片生成失败:确认指定模型存在、验证ComfyUI连接、尝试重启",
helpTrouble5: "LoRA Stack出现「No such file or directory」错误:该LoRA可能未被comfyui-lora-manager识别(已知问题)。请尝试其他LoRA,或检查该文件是否显示在Lora Manager自身的界面中",
helpTrouble6: "移动或删除模型后出现模型加载错误:ComfyUI会缓存文件路径,可能仍引用旧位置。请通过ComfyUI菜单执行Hard Reset(Queue → Hard Reset)或重启ComfyUI以清除缓存。",
helpTrouble7: "Tagger标签不显示模型:确认ComfyUI/models/tagger/<模型名>/文件夹存在且包含.onnx或.h5文件;如未安装onnxruntime,请运行:python_embeded\\python.exe -m pip install onnxruntime",
helpSupportTitle: "支持",
helpSupportDesc: "如果您觉得这个插件有用,请考虑支持它的开发!",
helpGithubDesc: "报告问题、请求功能和贡献代码",
+667
View File
@@ -0,0 +1,667 @@
/**
* Tagger Tab
* Handles WD Tagger / DeepDanbooru + Ollama VLM single/batch tagging.
*/
import { t } from "./i18n.js";
import { showToast } from "./app.js";
// ── 状態 ────────────────────────────────────────────────────────
const state = {
/** @type {string|null} 現在の画像 base64 (data URI prefix を除く) */
imageB64: null,
/** @type {string} 元のファイルパス (Galleryから来た場合) */
imagePath: "",
/** @type {string} 表示名 */
imageName: "",
/** @type {NodeJS.Timeout|null} バッチポーリングタイマー */
batchPollTimer: null,
/** @type {number|null} DB編集中の行ID */
dbEditId: null,
};
// ── 初期化 ──────────────────────────────────────────────────────
export function initTaggerTab() {
_applyI18n();
_setupSubtabs();
_setupSliders();
_setupOllamaToggle();
_setupSingleActions();
_setupBatchActions();
_setupDbActions();
_loadModels();
_loadSettings();
}
/** GalleryタブからTaggerタブへ画像を渡す */
export function openImageInTaggerTab(img) {
// タブを切り替え
document.querySelectorAll(".wfm-tab").forEach(b => b.classList.remove("active"));
document.querySelectorAll(".wfm-tab-content").forEach(s => s.classList.remove("active"));
const btn = document.querySelector('.wfm-tab[data-tab="tagger"]');
const sec = document.getElementById("wfm-tab-tagger");
if (btn) btn.classList.add("active");
if (sec) sec.classList.add("active");
// 画像を読み込む (Gallery APIから取得)
if (!img?.path) return;
const url = `/wfm/gallery/image/serve?path=${encodeURIComponent(img.path)}`;
fetch(url)
.then(r => {
if (!r.ok) throw new Error(`HTTP ${r.status}`);
return r.blob();
})
.then(blob => {
const reader = new FileReader();
reader.onload = e => {
_setPreviewFromDataUrl(e.target.result, img.filename || img.path.split(/[\\/]/).pop());
state.imagePath = img.path;
};
reader.readAsDataURL(blob);
})
.catch(() => showToast(t("taggerNoImage"), "error"));
}
// ── i18n ─────────────────────────────────────────────────────────
function _applyI18n() {
const map = {
"wfm-tagger-from-gallery-btn": "taggerFromGallery",
"wfm-tagger-upload-label": "taggerUpload",
"wfm-tagger-threshold-label": "taggerThreshold",
"wfm-tagger-char-threshold-label": "taggerCharThreshold",
"wfm-tagger-generate-btn": "taggerGenerateTags",
"wfm-tagger-ollama-header-label": "taggerOllamaEnable",
"wfm-tagger-ollama-url-label": "taggerOllamaUrl",
"wfm-tagger-ollama-model-label": "taggerOllamaModel",
"wfm-tagger-ollama-prompt-label": "taggerOllamaPrompt",
"wfm-tagger-ollama-max-label": "taggerOllamaMaxTags",
"wfm-tagger-result-header": "taggerResult",
"wfm-tagger-send-genui-btn": "taggerSendToGenUI",
"wfm-tagger-send-prompt-btn": "taggerSendToPrompt",
"wfm-tagger-save-gallery-btn": "taggerSaveToGallery",
"wfm-tagger-write-file-btn": "taggerWriteToFile",
"wfm-tagger-save-db-btn": "taggerSaveToDB",
"wfm-tagger-batch-folder-header": "taggerBatchFolder",
"wfm-tagger-batch-threshold-label": "taggerThreshold",
"wfm-tagger-batch-char-threshold-label": "taggerCharThreshold",
"wfm-tagger-batch-url-label": "taggerOllamaUrl",
"wfm-tagger-batch-model-label": "taggerOllamaModel",
"wfm-tagger-batch-prompt-label": "taggerOllamaPrompt",
"wfm-tagger-batch-output-header": "taggerBatchOutput",
"wfm-tagger-batch-save-db-label": "taggerSaveToDB",
"wfm-tagger-batch-write-file-label": "taggerWriteToFile",
"wfm-tagger-batch-write-txt-label": "taggerBatchWriteTxt",
"wfm-tagger-batch-start-btn": "taggerBatchStart",
"wfm-tagger-batch-stop-btn": "taggerBatchStop",
"wfm-tagger-db-search-btn": "taggerDbSearch",
"wfm-tagger-db-show-all-btn": "taggerDbShowAll",
"wfm-tagger-db-export-btn": "taggerDbExport",
"wfm-tagger-db-th-file": "taggerDbFile",
"wfm-tagger-db-th-interrogator": "taggerDbInterrogator",
"wfm-tagger-db-th-vlm": "taggerDbVlm",
"wfm-tagger-db-th-date": "taggerDbDate",
"wfm-tagger-db-edit-interr-label": "taggerDbInterrogator",
"wfm-tagger-db-edit-vlm-label": "taggerDbVlm",
"wfm-tagger-db-save-btn": "taggerDbSave",
"wfm-tagger-db-delete-btn": "taggerDbDelete",
};
for (const [id, key] of Object.entries(map)) {
const el = document.getElementById(id);
if (el) el.textContent = t(key);
}
// サブタブボタン
const subtabKeys = ["taggerSingleTab", "taggerBatchTab", "taggerDbTab"];
document.querySelectorAll(".wfm-tagger-subtab-btn").forEach((btn, i) => {
btn.textContent = t(subtabKeys[i] || "");
});
// プレースホルダー
const ph = document.getElementById("wfm-tagger-preview-placeholder");
if (ph) ph.textContent = t("taggerNoImage");
}
// ── サブタブ切り替え ─────────────────────────────────────────────
function _setupSubtabs() {
document.querySelectorAll(".wfm-tagger-subtab-btn").forEach(btn => {
btn.addEventListener("click", () => {
document.querySelectorAll(".wfm-tagger-subtab-btn").forEach(b => b.classList.remove("active"));
btn.classList.add("active");
const tab = btn.dataset.taggerTab;
document.querySelectorAll(".wfm-tagger-subtab").forEach(s => {
s.style.display = s.id === `wfm-tagger-${tab}` ? "" : "none";
});
if (tab === "db") _dbLoad();
});
});
}
// ── スライダー ───────────────────────────────────────────────────
function _setupSliders() {
[
["wfm-tagger-threshold", "wfm-tagger-threshold-val"],
["wfm-tagger-char-threshold", "wfm-tagger-char-threshold-val"],
["wfm-tagger-batch-threshold", "wfm-tagger-batch-threshold-val"],
["wfm-tagger-batch-char-threshold", "wfm-tagger-batch-char-threshold-val"],
].forEach(([sliderId, valId]) => {
const slider = document.getElementById(sliderId);
const val = document.getElementById(valId);
if (!slider || !val) return;
val.textContent = slider.value;
slider.addEventListener("input", () => { val.textContent = slider.value; });
});
}
// ── Ollama トグル ────────────────────────────────────────────────
function _setupOllamaToggle() {
const chk = document.getElementById("wfm-tagger-ollama-enable");
const wrap = document.getElementById("wfm-tagger-ollama-settings");
if (chk && wrap) {
chk.addEventListener("change", () => {
wrap.style.display = chk.checked ? "" : "none";
});
wrap.style.display = chk.checked ? "" : "none";
}
}
// ── モデル一覧ロード ─────────────────────────────────────────────
async function _loadModels() {
try {
const res = await fetch("/wfm/tagger/models");
const data = await res.json();
const models = data.models || [];
["wfm-tagger-model", "wfm-tagger-batch-model"].forEach(selId => {
const sel = document.getElementById(selId);
if (!sel) return;
sel.innerHTML = `<option value="">${t("taggerNoModel")}</option>` +
models.map(m => `<option value="${m.name}">[${m.type}] ${m.name}</option>`).join("");
});
} catch (e) {
console.warn("tagger: _loadModels error", e);
}
}
// ── Ollamaモデル一覧ロード ───────────────────────────────────────
async function _loadOllamaModels(urlId, selId) {
const urlEl = document.getElementById(urlId);
const selEl = document.getElementById(selId);
if (!urlEl || !selEl) return;
const api = urlEl.value || "http://127.0.0.1:11434";
try {
const res = await fetch(`/wfm/tagger/ollama/models?api_url=${encodeURIComponent(api)}`);
const data = await res.json();
const models = data.models || [];
if (models.length === 0) {
selEl.innerHTML = `<option value="">(${t("taggerNoModel")})</option>`;
} else {
selEl.innerHTML = models.map(m => `<option value="${m}">${m}</option>`).join("");
}
} catch {
selEl.innerHTML = `<option value="">(${t("taggerNoModel")})</option>`;
}
}
// ── 設定の読み込み/保存 ─────────────────────────────────────────
async function _loadSettings() {
try {
const res = await fetch("/wfm/tagger/settings");
const data = await res.json();
if (data.ollama_url) {
const el = document.getElementById("wfm-tagger-ollama-url");
if (el) el.value = data.ollama_url;
}
if (data.batch_ollama_url) {
const el = document.getElementById("wfm-tagger-batch-ollama-url");
if (el) el.value = data.batch_ollama_url;
}
} catch {
// 初回は無視
}
}
async function _saveSettings() {
try {
await fetch("/wfm/tagger/settings", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({
ollama_url: document.getElementById("wfm-tagger-ollama-url")?.value || "",
batch_ollama_url: document.getElementById("wfm-tagger-batch-ollama-url")?.value || "",
}),
});
} catch {
// サイレントに失敗
}
}
// ── Single サブタブ ─────────────────────────────────────────────
function _setupSingleActions() {
// ファイルアップロード
document.getElementById("wfm-tagger-upload")?.addEventListener("change", e => {
const file = e.target.files?.[0];
if (!file) return;
_loadImageFile(file);
e.target.value = "";
});
// ドラッグ&ドロップ
const dropZone = document.getElementById("wfm-tagger-drop-zone");
if (dropZone) {
dropZone.addEventListener("dragover", e => {
e.preventDefault();
dropZone.classList.add("drag-over");
});
dropZone.addEventListener("dragleave", e => {
if (!dropZone.contains(e.relatedTarget)) {
dropZone.classList.remove("drag-over");
}
});
dropZone.addEventListener("drop", e => {
e.preventDefault();
dropZone.classList.remove("drag-over");
const file = e.dataTransfer.files?.[0];
if (file && file.type.startsWith("image/")) {
_loadImageFile(file);
}
});
}
// タグ生成
document.getElementById("wfm-tagger-generate-btn")?.addEventListener("click", _generateTags);
// Ollamaリフレッシュ
document.getElementById("wfm-tagger-ollama-refresh-btn")?.addEventListener("click", () => {
_loadOllamaModels("wfm-tagger-ollama-url", "wfm-tagger-ollama-model");
});
// 結果アクション
document.getElementById("wfm-tagger-send-genui-btn")?.addEventListener("click", _sendToGenUI);
document.getElementById("wfm-tagger-send-prompt-btn")?.addEventListener("click", _sendToPrompt);
document.getElementById("wfm-tagger-save-gallery-btn")?.addEventListener("click", _saveToGallery);
document.getElementById("wfm-tagger-write-file-btn")?.addEventListener("click", _writeToFile);
document.getElementById("wfm-tagger-save-db-btn")?.addEventListener("click", _saveToDb);
}
function _loadImageFile(file) {
const reader = new FileReader();
reader.onload = e => {
_setPreviewFromDataUrl(e.target.result, file.name);
state.imagePath = "";
};
reader.readAsDataURL(file);
}
function _setPreviewFromDataUrl(dataUrl, name) {
state.imageB64 = dataUrl.replace(/^data:[^;]+;base64,/, "");
state.imageName = name;
const img = document.getElementById("wfm-tagger-preview");
const ph = document.getElementById("wfm-tagger-preview-placeholder");
const nameEl = document.getElementById("wfm-tagger-image-name");
if (img) { img.src = dataUrl; img.style.display = ""; }
if (ph) ph.style.display = "none";
if (nameEl) nameEl.textContent = name;
}
async function _generateTags() {
if (!state.imageB64) {
showToast(t("taggerNoImage"), "warning");
return;
}
const model = document.getElementById("wfm-tagger-model")?.value || "";
const threshold = parseFloat(document.getElementById("wfm-tagger-threshold")?.value || "0.35");
const charThreshold = parseFloat(document.getElementById("wfm-tagger-char-threshold")?.value || "0.85");
const useOllama = document.getElementById("wfm-tagger-ollama-enable")?.checked || false;
const ollamaUrl = document.getElementById("wfm-tagger-ollama-url")?.value || "http://127.0.0.1:11434";
const ollamaModel = document.getElementById("wfm-tagger-ollama-model")?.value || "";
const ollamaPrompt = document.getElementById("wfm-tagger-ollama-prompt")?.value || "";
const maxTags = parseInt(document.getElementById("wfm-tagger-ollama-max-tags")?.value || "40");
const generateBtn = document.getElementById("wfm-tagger-generate-btn");
if (generateBtn) { generateBtn.disabled = true; generateBtn.textContent = "..."; }
try {
let tags = "";
if (model) {
const res = await fetch("/wfm/tagger/predict", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ image_b64: state.imageB64, model, threshold, char_threshold: charThreshold }),
});
const data = await res.json();
if (data.error) throw new Error(data.error);
tags = data.tags || "";
}
if (useOllama && ollamaModel) {
const res2 = await fetch("/wfm/tagger/ollama/predict", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({
image_b64: state.imageB64, api_url: ollamaUrl,
model: ollamaModel, prompt: ollamaPrompt, max_tags: maxTags,
}),
});
const data2 = await res2.json();
if (data2.error) throw new Error(data2.error);
const vlm = data2.tags || "";
tags = [tags, vlm].filter(Boolean).join(", ");
}
if (!tags) { showToast(t("taggerNoTags"), "warning"); return; }
document.getElementById("wfm-tagger-result").value = tags;
showToast(t("taggerDone"), "success");
_saveSettings();
} catch (e) {
showToast(e.message, "error");
} finally {
if (generateBtn) { generateBtn.disabled = false; generateBtn.textContent = t("taggerGenerateTags"); }
}
}
function _sendToGenUI() {
const tags = document.getElementById("wfm-tagger-result")?.value || "";
if (!tags) { showToast(t("taggerNoTags"), "warning"); return; }
const ta = document.getElementById("wfm-prompt-pos-text");
if (!ta) { showToast(t("taggerNoGenUI"), "warning"); return; }
const cur = ta.value;
ta.value = cur ? `${cur}, ${tags}` : tags;
ta.dispatchEvent(new Event("input", { bubbles: true }));
document.getElementById("wfm-prompt-pos-apply")?.click();
showToast(t("taggerSentToGenUI"), "success");
}
function _sendToPrompt() {
const tags = document.getElementById("wfm-tagger-result")?.value || "";
if (!tags) { showToast(t("taggerNoTags"), "warning"); return; }
const ta = document.getElementById("wfm-preset-pos");
if (!ta) { showToast(t("taggerNoGalleryImage"), "warning"); return; }
const cur = ta.value;
ta.value = cur ? `${cur}, ${tags}` : tags;
ta.dispatchEvent(new Event("input", { bubbles: true }));
showToast(t("taggerSentToPrompt"), "success");
}
async function _saveToGallery() {
const tags = document.getElementById("wfm-tagger-result")?.value || "";
if (!tags || !state.imagePath) {
showToast(t("taggerNoGalleryImage"), "warning");
return;
}
try {
const tagsArray = tags.split(",").map(s => s.trim()).filter(Boolean);
const res = await fetch("/wfm/gallery/image/meta", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ path: state.imagePath, tags: tagsArray }),
});
const data = await res.json();
if (data.error) throw new Error(data.error);
showToast(t("taggerSavedToGallery"), "success");
} catch (e) {
showToast(e.message, "error");
}
}
async function _writeToFile() {
const tags = document.getElementById("wfm-tagger-result")?.value || "";
if (!tags || !state.imagePath) {
showToast(t("taggerNoGalleryImage"), "warning");
return;
}
try {
const res = await fetch("/wfm/tagger/write_meta", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ path: state.imagePath, tags }),
});
const data = await res.json();
if (data.error) throw new Error(data.error);
showToast(t("taggerWrittenToFile"), "success");
} catch (e) {
showToast(e.message, "error");
}
}
async function _saveToDb() {
const tags = document.getElementById("wfm-tagger-result")?.value || "";
if (!tags) { showToast(t("taggerNoTags"), "warning"); return; }
try {
const res = await fetch("/wfm/tagger/db/save", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ filename: state.imageName, interrogator_tags: tags, vlm_tags: "" }),
});
const data = await res.json();
if (data.error) throw new Error(data.error);
showToast(t("taggerSavedToDB"), "success");
} catch (e) {
showToast(e.message, "error");
}
}
// ── Batch サブタブ ───────────────────────────────────────────────
function _setupBatchActions() {
document.getElementById("wfm-tagger-batch-ollama-refresh-btn")?.addEventListener("click", () => {
_loadOllamaModels("wfm-tagger-batch-ollama-url", "wfm-tagger-batch-ollama-model");
});
document.getElementById("wfm-tagger-batch-start-btn")?.addEventListener("click", _batchStart);
document.getElementById("wfm-tagger-batch-stop-btn")?.addEventListener("click", _batchStop);
}
async function _batchStart() {
const folder = document.getElementById("wfm-tagger-batch-folder")?.value || "";
if (!folder) { showToast(t("taggerBatchFolder"), "warning"); return; }
const body = {
folder,
model: document.getElementById("wfm-tagger-batch-model")?.value || "",
threshold: parseFloat(document.getElementById("wfm-tagger-batch-threshold")?.value || "0.35"),
char_threshold: parseFloat(document.getElementById("wfm-tagger-batch-char-threshold")?.value || "0.85"),
use_ollama: document.getElementById("wfm-tagger-batch-ollama-enable")?.checked || false,
ollama_api: document.getElementById("wfm-tagger-batch-ollama-url")?.value || "http://127.0.0.1:11434",
ollama_model: document.getElementById("wfm-tagger-batch-ollama-model")?.value || "",
ollama_prompt: document.getElementById("wfm-tagger-batch-ollama-prompt")?.value || "",
save_db: document.getElementById("wfm-tagger-batch-save-db")?.checked ?? true,
write_file: document.getElementById("wfm-tagger-batch-write-file")?.checked || false,
write_txt: document.getElementById("wfm-tagger-batch-write-txt")?.checked || false,
};
try {
const res = await fetch("/wfm/tagger/batch/start", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify(body),
});
const data = await res.json();
if (data.error) throw new Error(data.error);
document.getElementById("wfm-tagger-batch-start-btn").disabled = true;
document.getElementById("wfm-tagger-batch-stop-btn").disabled = false;
document.getElementById("wfm-tagger-batch-log").value = "";
_startBatchPoll();
} catch (e) {
showToast(e.message, "error");
}
}
async function _batchStop() {
await fetch("/wfm/tagger/batch/stop", { method: "POST" });
}
function _startBatchPoll() {
if (state.batchPollTimer) clearInterval(state.batchPollTimer);
state.batchPollTimer = setInterval(_pollBatch, 1000);
}
async function _pollBatch() {
try {
const res = await fetch("/wfm/tagger/batch/status");
const data = await res.json();
const total = data.total || 0;
const done = data.done || 0;
const pct = total > 0 ? Math.round((done / total) * 100) : 0;
const bar = document.getElementById("wfm-tagger-batch-progress-bar");
const txt = document.getElementById("wfm-tagger-batch-progress-text");
if (bar) bar.style.width = pct + "%";
if (txt) txt.textContent = `${done} / ${total} (${pct}%)`;
const logEl = document.getElementById("wfm-tagger-batch-log");
if (logEl && data.log?.length) {
logEl.value = data.log.join("\n");
logEl.scrollTop = logEl.scrollHeight;
}
if (!data.running) {
clearInterval(state.batchPollTimer);
state.batchPollTimer = null;
document.getElementById("wfm-tagger-batch-start-btn").disabled = false;
document.getElementById("wfm-tagger-batch-stop-btn").disabled = true;
showToast(t("taggerDone"), "success");
}
} catch {
// 無視
}
}
// ── DB サブタブ ──────────────────────────────────────────────────
function _setupDbActions() {
document.getElementById("wfm-tagger-db-search-btn")?.addEventListener("click", _dbSearch);
document.getElementById("wfm-tagger-db-show-all-btn")?.addEventListener("click", _dbLoad);
document.getElementById("wfm-tagger-db-export-btn")?.addEventListener("click", _dbExport);
document.getElementById("wfm-tagger-db-search-input")?.addEventListener("keydown", e => {
if (e.key === "Enter") _dbSearch();
});
document.getElementById("wfm-tagger-db-save-btn")?.addEventListener("click", _dbSave);
document.getElementById("wfm-tagger-db-delete-btn")?.addEventListener("click", _dbDelete);
}
async function _dbLoad() {
try {
const res = await fetch("/wfm/tagger/db/list?limit=200&offset=0");
const data = await res.json();
_renderDbTable(data.rows || []);
} catch (e) {
showToast(e.message, "error");
}
}
async function _dbSearch() {
const q = document.getElementById("wfm-tagger-db-search-input")?.value || "";
if (!q) { _dbLoad(); return; }
try {
const res = await fetch(`/wfm/tagger/db/search?q=${encodeURIComponent(q)}`);
const data = await res.json();
_renderDbTable(data.rows || []);
} catch (e) {
showToast(e.message, "error");
}
}
function _renderDbTable(rows) {
const tbody = document.getElementById("wfm-tagger-db-tbody");
if (!tbody) return;
if (!rows.length) {
tbody.innerHTML = `<tr><td colspan="5" style="text-align:center;padding:20px;color:var(--wfm-text-secondary)">${t("taggerDbEmpty")}</td></tr>`;
return;
}
tbody.innerHTML = rows.map(row => `
<tr data-id="${row.id}" title="${_esc(row.all_tags)}">
<td>${row.id}</td>
<td>${_esc(row.filename)}</td>
<td>${_esc(_truncate(row.interrogator_tags, 60))}</td>
<td>${_esc(_truncate(row.vlm_tags, 60))}</td>
<td>${_esc(row.created_at?.split("T")[0] || row.created_at || "")}</td>
</tr>
`).join("");
tbody.querySelectorAll("tr").forEach(tr => {
tr.addEventListener("click", () => {
tbody.querySelectorAll("tr").forEach(r => r.classList.remove("selected"));
tr.classList.add("selected");
const row = rows.find(r => r.id === parseInt(tr.dataset.id));
if (row) _openDbEdit(row);
});
});
}
function _openDbEdit(row) {
state.dbEditId = row.id;
const panel = document.getElementById("wfm-tagger-db-edit-panel");
const header = document.getElementById("wfm-tagger-db-edit-header");
const interrEl = document.getElementById("wfm-tagger-db-edit-interrogator");
const vlmEl = document.getElementById("wfm-tagger-db-edit-vlm");
if (!panel) return;
if (header) header.textContent = `ID: ${row.id} ${row.filename}`;
if (interrEl) interrEl.value = row.interrogator_tags;
if (vlmEl) vlmEl.value = row.vlm_tags;
panel.style.display = "";
}
async function _dbSave() {
if (state.dbEditId == null) return;
const interrogator_tags = document.getElementById("wfm-tagger-db-edit-interrogator")?.value || "";
const vlm_tags = document.getElementById("wfm-tagger-db-edit-vlm")?.value || "";
try {
const res = await fetch(`/wfm/tagger/db/${state.dbEditId}`, {
method: "PUT",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ interrogator_tags, vlm_tags }),
});
const data = await res.json();
if (data.error) throw new Error(data.error);
showToast(t("taggerDbSaved"), "success");
_dbLoad();
} catch (e) {
showToast(e.message, "error");
}
}
async function _dbDelete() {
if (state.dbEditId == null) return;
if (!confirm(`Delete DB record #${state.dbEditId}?`)) return;
try {
const res = await fetch(`/wfm/tagger/db/${state.dbEditId}`, { method: "DELETE" });
const data = await res.json();
if (data.error) throw new Error(data.error);
document.getElementById("wfm-tagger-db-edit-panel").style.display = "none";
state.dbEditId = null;
showToast(t("taggerDbDeleted"), "success");
_dbLoad();
} catch (e) {
showToast(e.message, "error");
}
}
function _dbExport() {
const a = document.createElement("a");
a.href = "/wfm/tagger/db/export";
a.download = "tagger_tags.csv";
a.click();
}
// ── ユーティリティ ───────────────────────────────────────────────
function _esc(str) {
return String(str || "").replace(/&/g, "&amp;").replace(/</g, "&lt;").replace(/>/g, "&gt;").replace(/"/g, "&quot;");
}
function _truncate(str, len) {
if (!str) return "";
return str.length > len ? str.slice(0, len) + "…" : str;
}
+245
View File
@@ -7,6 +7,7 @@
<link rel="icon" type="image/svg+xml" href="/wfm_static/favicon.svg">
<link rel="stylesheet" href="/wfm_static/css/main.css">
<link rel="stylesheet" href="/wfm_static/css/gallery-tab.css">
<link rel="stylesheet" href="/wfm_static/css/tagger-tab.css">
</head>
<body>
<!-- Header -->
@@ -22,6 +23,7 @@
<button class="wfm-tab" data-tab="prompt">Prompt</button>
<button class="wfm-tab" data-tab="metadata">Metadata</button>
<button class="wfm-tab" data-tab="gallery">Gallery</button>
<button class="wfm-tab" data-tab="tagger">Tagger</button>
<button class="wfm-tab" data-tab="settings">Settings</button>
<button class="wfm-tab" data-tab="help">Help</button>
<button class="wfm-tab" data-tab="ai">AI TOOL</button>
@@ -920,6 +922,7 @@
<button class="wfm-help-nav-item" data-help-page="models">Models Tab</button>
<button class="wfm-help-nav-item" data-help-page="sidepanel">Library (Side Panel)</button>
<button class="wfm-help-nav-item" data-help-page="aitool">AI TOOL Tab</button>
<button class="wfm-help-nav-item" data-help-page="tagger">Tagger Tab</button>
<button class="wfm-help-nav-item" data-help-page="shortcuts">Keyboard Shortcuts</button>
<button class="wfm-help-nav-item" data-help-page="trouble">Troubleshooting</button>
</div>
@@ -1080,6 +1083,7 @@
<li id="wfm-help-gallery-11">Supports PNG (tEXt/iTXt), JPEG, WebP, GIF; workflow extraction reads ComfyUI prompt and workflow keys embedded in PNG metadata</li>
<li id="wfm-help-gallery-12"><b>Metadata</b> button (green, in the detail panel tab bar): opens the selected image in the Metadata tab for full embedded-data inspection; Alt+click on any image thumbnail has the same effect</li>
<li id="wfm-help-gallery-13"><b>Load GenUI</b> button (blue, in the detail panel tab bar): loads the embedded ComfyUI workflow from the selected image directly into the GenerateUI tab and switches to it; shows a warning toast if no workflow is embedded or the format is not supported</li>
<li id="wfm-help-gallery-14"><b>Tagger</b> button (purple, in the detail panel action row): opens the selected image directly in the Tagger tab's Single mode for instant tagging</li>
</ul>
</div>
</div>
@@ -1141,6 +1145,23 @@
</div>
</div>
<div class="wfm-help-page" id="wfm-help-page-tagger">
<div class="wfm-help-card">
<h3 id="wfm-help-tagger-title">Tagger Tab</h3>
<ul>
<li id="wfm-help-tagger-1"></li>
<li id="wfm-help-tagger-2"></li>
<li id="wfm-help-tagger-3"></li>
<li id="wfm-help-tagger-4"></li>
<li id="wfm-help-tagger-5"></li>
<li id="wfm-help-tagger-6"></li>
<li id="wfm-help-tagger-7"></li>
<li id="wfm-help-tagger-8"></li>
<li id="wfm-help-tagger-9"></li>
</ul>
</div>
</div>
<div class="wfm-help-page" id="wfm-help-page-shortcuts">
<div class="wfm-help-card">
<h3 id="wfm-help-shortcuts-title">Keyboard Shortcuts</h3>
@@ -1163,6 +1184,7 @@
<li id="wfm-help-trouble-4">Image generation fails: Check if specified models exist in ComfyUI, verify connection, try restarting</li>
<li id="wfm-help-trouble-5">LoRA Stack fails with "No such file or directory" error: The LoRA may not be recognized by comfyui-lora-manager (known issue). Try a different LoRA or check if the file appears in Lora Manager's own UI</li>
<li id="wfm-help-trouble-6">Model loading error after moving or deleting models: ComfyUI caches file paths and may still reference old locations. Perform a Hard Reset from the ComfyUI menu (Queue → Hard Reset) or restart ComfyUI to clear the cache.</li>
<li id="wfm-help-trouble-7"></li>
</ul>
</div>
</div>
@@ -1274,8 +1296,11 @@
<button class="wfm-gallery-detail-tab-btn active" data-detail-tab="info">Info</button>
<button class="wfm-gallery-detail-tab-btn" data-detail-tab="meta">JSON</button>
<button class="wfm-gallery-detail-tab-btn" data-detail-tab="group">Groups</button>
</div>
<div class="wfm-gallery-detail-action-row">
<button id="wfm-gallery-open-metadata-btn" class="wfm-gallery-action-btn wfm-gallery-action-btn-green" title="Open in Metadata tab (or Alt+click image)">Metadata</button>
<button id="wfm-gallery-load-genui-btn" class="wfm-gallery-action-btn wfm-gallery-action-btn-primary" title="Load workflow into GenerateUI">Load GenUI</button>
<button id="wfm-gallery-open-tagger-btn" class="wfm-gallery-action-btn wfm-gallery-action-btn-tagger" title="Open in Tagger"></button>
</div>
<!-- Info タブ -->
<div id="wfm-gallery-detail-info" class="wfm-gallery-detail-tab-content active">
@@ -1325,6 +1350,226 @@
</div>
</section>
<!-- Tagger Tab -->
<section id="wfm-tab-tagger" class="wfm-tab-content">
<div class="wfm-tagger-subtab-nav">
<button class="wfm-tagger-subtab-btn active" data-tagger-tab="single"></button>
<button class="wfm-tagger-subtab-btn" data-tagger-tab="batch"></button>
<button class="wfm-tagger-subtab-btn" data-tagger-tab="db"></button>
</div>
<!-- Single サブタブ -->
<div id="wfm-tagger-single" class="wfm-tagger-subtab">
<div class="wfm-tagger-single-layout">
<!-- 左: 画像 -->
<div class="wfm-tagger-image-pane">
<div class="wfm-tagger-image-sources">
<label class="wfm-btn wfm-btn-sm" style="cursor:pointer">
<span id="wfm-tagger-upload-label"></span>
<input type="file" id="wfm-tagger-upload" accept="image/*" style="display:none">
</label>
</div>
<div class="wfm-tagger-preview-wrap" id="wfm-tagger-drop-zone">
<img id="wfm-tagger-preview" style="display:none; max-width:100%; max-height:100%; object-fit:contain;">
<div id="wfm-tagger-preview-placeholder" class="wfm-placeholder"></div>
</div>
<div id="wfm-tagger-image-name" class="wfm-tagger-image-name"></div>
</div>
<!-- 右: 設定・結果 -->
<div class="wfm-tagger-controls-pane">
<!-- WD Tagger -->
<div class="wfm-tagger-section">
<div class="wfm-tagger-section-header">WD Tagger / DeepDanbooru</div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label"></label>
<select id="wfm-tagger-model" class="wfm-select wfm-tagger-select"></select>
</div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-threshold-label"></label>
<input type="range" id="wfm-tagger-threshold" min="0" max="1" step="0.01" value="0.35" class="wfm-tagger-slider">
<span id="wfm-tagger-threshold-val" class="wfm-tagger-slider-val">0.35</span>
</div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-char-threshold-label"></label>
<input type="range" id="wfm-tagger-char-threshold" min="0" max="1" step="0.01" value="0.85" class="wfm-tagger-slider">
<span id="wfm-tagger-char-threshold-val" class="wfm-tagger-slider-val">0.85</span>
</div>
<button id="wfm-tagger-generate-btn" class="wfm-btn wfm-btn-primary" style="width:100%;margin-top:6px;"></button>
</div>
<!-- Ollama VLM -->
<div class="wfm-tagger-section">
<div class="wfm-tagger-section-header">
<label style="display:flex;align-items:center;gap:6px;cursor:pointer;font-weight:600;">
<input type="checkbox" id="wfm-tagger-ollama-enable">
<span id="wfm-tagger-ollama-header-label">Ollama VLM</span>
</label>
</div>
<div id="wfm-tagger-ollama-settings" style="display:none">
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-ollama-url-label"></label>
<input type="text" id="wfm-tagger-ollama-url" class="wfm-search-input" value="http://127.0.0.1:11434" style="flex:1;">
</div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-ollama-model-label"></label>
<select id="wfm-tagger-ollama-model" class="wfm-select wfm-tagger-select"></select>
<button id="wfm-tagger-ollama-refresh-btn" class="wfm-btn wfm-btn-sm" title="Refresh models">↻</button>
</div>
<div class="wfm-form-row wfm-tagger-form-row" style="align-items:flex-start">
<label class="wfm-tagger-label" id="wfm-tagger-ollama-prompt-label" style="padding-top:4px;"></label>
<textarea id="wfm-tagger-ollama-prompt" class="wfm-textarea" rows="3" style="flex:1;resize:vertical;">Output danbooru-style, comma-separated, English tags for this image. ONLY output the tags. Do not use double quotes or hyphens in the tags.</textarea>
</div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-ollama-max-label"></label>
<input type="number" id="wfm-tagger-ollama-max-tags" class="wfm-search-input" value="40" min="1" max="500" style="width:70px;">
</div>
</div>
</div>
<!-- 結果 -->
<div class="wfm-tagger-section">
<div class="wfm-tagger-section-header" id="wfm-tagger-result-header"></div>
<textarea id="wfm-tagger-result" class="wfm-textarea" rows="5" style="width:100%;resize:vertical;" placeholder="Tags will appear here..."></textarea>
<div class="wfm-tagger-result-actions">
<button id="wfm-tagger-send-genui-btn" class="wfm-btn wfm-btn-sm"></button>
<button id="wfm-tagger-send-prompt-btn" class="wfm-btn wfm-btn-sm"></button>
<button id="wfm-tagger-save-gallery-btn" class="wfm-btn wfm-btn-sm"></button>
<button id="wfm-tagger-write-file-btn" class="wfm-btn wfm-btn-sm"></button>
<button id="wfm-tagger-save-db-btn" class="wfm-btn wfm-btn-sm"></button>
</div>
</div>
</div>
</div>
</div>
<!-- Batch サブタブ -->
<div id="wfm-tagger-batch" class="wfm-tagger-subtab" style="display:none">
<div class="wfm-tagger-batch-layout">
<div class="wfm-tagger-section">
<div class="wfm-tagger-section-header" id="wfm-tagger-batch-folder-header"></div>
<div class="wfm-form-row wfm-tagger-form-row">
<input type="text" id="wfm-tagger-batch-folder" class="wfm-search-input" style="flex:1;" placeholder="C:\path\to\images">
</div>
</div>
<div class="wfm-tagger-batch-cols">
<!-- WD Tagger 設定 -->
<div class="wfm-tagger-section">
<div class="wfm-tagger-section-header">WD Tagger / DeepDanbooru</div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label"></label>
<select id="wfm-tagger-batch-model" class="wfm-select wfm-tagger-select"></select>
</div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-batch-threshold-label"></label>
<input type="range" id="wfm-tagger-batch-threshold" min="0" max="1" step="0.01" value="0.35" class="wfm-tagger-slider">
<span id="wfm-tagger-batch-threshold-val" class="wfm-tagger-slider-val">0.35</span>
</div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-batch-char-threshold-label"></label>
<input type="range" id="wfm-tagger-batch-char-threshold" min="0" max="1" step="0.01" value="0.85" class="wfm-tagger-slider">
<span id="wfm-tagger-batch-char-threshold-val" class="wfm-tagger-slider-val">0.85</span>
</div>
</div>
<!-- Ollama 設定 -->
<div class="wfm-tagger-section">
<div class="wfm-tagger-section-header">
<label style="display:flex;align-items:center;gap:6px;cursor:pointer;font-weight:600;">
<input type="checkbox" id="wfm-tagger-batch-ollama-enable">
Ollama VLM
</label>
</div>
<div id="wfm-tagger-batch-ollama-settings">
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-batch-url-label"></label>
<input type="text" id="wfm-tagger-batch-ollama-url" class="wfm-search-input" value="http://127.0.0.1:11434" style="flex:1;">
</div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-batch-model-label"></label>
<select id="wfm-tagger-batch-ollama-model" class="wfm-select wfm-tagger-select"></select>
<button id="wfm-tagger-batch-ollama-refresh-btn" class="wfm-btn wfm-btn-sm">↻</button>
</div>
<div class="wfm-form-row wfm-tagger-form-row" style="align-items:flex-start">
<label class="wfm-tagger-label" id="wfm-tagger-batch-prompt-label" style="padding-top:4px;"></label>
<textarea id="wfm-tagger-batch-ollama-prompt" class="wfm-textarea" rows="3" style="flex:1;resize:vertical;">Output danbooru-style, comma-separated, English tags for this image. ONLY output the tags.</textarea>
</div>
</div>
</div>
</div>
<!-- 出力設定 -->
<div class="wfm-tagger-section">
<div class="wfm-tagger-section-header" id="wfm-tagger-batch-output-header"></div>
<div style="display:flex;gap:16px;">
<label style="display:flex;align-items:center;gap:6px;cursor:pointer;">
<input type="checkbox" id="wfm-tagger-batch-save-db" checked>
<span id="wfm-tagger-batch-save-db-label"></span>
</label>
<label style="display:flex;align-items:center;gap:6px;cursor:pointer;">
<input type="checkbox" id="wfm-tagger-batch-write-file">
<span id="wfm-tagger-batch-write-file-label"></span>
</label>
<label style="display:flex;align-items:center;gap:6px;cursor:pointer;">
<input type="checkbox" id="wfm-tagger-batch-write-txt">
<span id="wfm-tagger-batch-write-txt-label"></span>
</label>
</div>
</div>
<!-- ボタン + プログレス -->
<div style="display:flex;gap:8px;align-items:center;margin-bottom:8px;">
<button id="wfm-tagger-batch-start-btn" class="wfm-btn wfm-btn-primary"></button>
<button id="wfm-tagger-batch-stop-btn" class="wfm-btn wfm-btn-danger" disabled></button>
<span id="wfm-tagger-batch-progress-text" style="font-size:12px;color:var(--wfm-text-muted);"></span>
</div>
<div class="wfm-tagger-progress-bar-wrap">
<div id="wfm-tagger-batch-progress-bar" class="wfm-tagger-progress-bar" style="width:0%"></div>
</div>
<textarea id="wfm-tagger-batch-log" class="wfm-textarea" rows="10" readonly style="width:100%;margin-top:8px;font-family:monospace;font-size:11px;"></textarea>
</div>
</div>
<!-- DB サブタブ -->
<div id="wfm-tagger-db" class="wfm-tagger-subtab" style="display:none">
<div class="wfm-tagger-db-layout">
<div class="wfm-tagger-db-toolbar">
<input type="text" id="wfm-tagger-db-search-input" class="wfm-search-input" style="flex:1;" placeholder="Search tags or filename...">
<button id="wfm-tagger-db-search-btn" class="wfm-btn wfm-btn-sm wfm-btn-primary"></button>
<button id="wfm-tagger-db-show-all-btn" class="wfm-btn wfm-btn-sm"></button>
<button id="wfm-tagger-db-export-btn" class="wfm-btn wfm-btn-sm"></button>
</div>
<div id="wfm-tagger-db-table-wrap" class="wfm-tagger-db-table-wrap">
<table id="wfm-tagger-db-table" class="wfm-table wfm-tagger-db-table">
<thead>
<tr>
<th>ID</th>
<th id="wfm-tagger-db-th-file"></th>
<th id="wfm-tagger-db-th-interrogator"></th>
<th id="wfm-tagger-db-th-vlm"></th>
<th id="wfm-tagger-db-th-date"></th>
</tr>
</thead>
<tbody id="wfm-tagger-db-tbody"></tbody>
</table>
</div>
<!-- 編集パネル -->
<div id="wfm-tagger-db-edit-panel" class="wfm-tagger-db-edit-panel" style="display:none">
<div class="wfm-tagger-section-header" id="wfm-tagger-db-edit-header"></div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-db-edit-interr-label"></label>
<textarea id="wfm-tagger-db-edit-interrogator" class="wfm-textarea" rows="3" style="flex:1;resize:vertical;"></textarea>
</div>
<div class="wfm-form-row wfm-tagger-form-row">
<label class="wfm-tagger-label" id="wfm-tagger-db-edit-vlm-label"></label>
<textarea id="wfm-tagger-db-edit-vlm" class="wfm-textarea" rows="3" style="flex:1;resize:vertical;"></textarea>
</div>
<div style="display:flex;gap:8px;margin-top:6px;">
<button id="wfm-tagger-db-save-btn" class="wfm-btn wfm-btn-sm wfm-btn-primary"></button>
<button id="wfm-tagger-db-delete-btn" class="wfm-btn wfm-btn-sm wfm-btn-danger"></button>
</div>
</div>
</div>
</div>
</section>
<!-- AI Tab -->
<section id="wfm-tab-ai" class="wfm-tab-content">
<!-- Translation pane -->